{"id":"https://openalex.org/W7214747136","doi":"https://doi.org/10.48550/arxiv.2609.34459","title":"Escaping Local Views: Discovering Latent Concepts for Interpretable Multi-Agent Reinforcement Learning","display_name":"Escaping Local Views: Discovering Latent Concepts for Interpretable Multi-Agent Reinforcement Learning","publication_year":2026,"publication_date":"2026-09-28","ids":{"openalex":"https://openalex.org/W7214747136","doi":"https://doi.org/10.48550/arxiv.2609.34459"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2609.34459","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2609.34459","pdf_url":"https://arxiv.org/pdf/2609.34459","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/2609.34459","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5153521969","display_name":"Yijie Sun","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I881766915","display_name":"Nanjing University","ror":"https://ror.org/01rxvg760","country_code":"CN","type":"education","lineage":["https://openalex.org/I881766915"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Sun, Yijie","raw_affiliation_strings":["Nanjing University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Nanjing University","institution_ids":["https://openalex.org/I881766915"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5153472728","display_name":"Sanquan Sun","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I14243506","display_name":"Hong Kong Polytechnic University","ror":"https://ror.org/0030zas98","country_code":"HK","type":"education","lineage":["https://openalex.org/I14243506"]}],"countries":["HK"],"is_corresponding":false,"raw_author_name":"Sun, Sanquan","raw_affiliation_strings":["Hong Kong Polytechnic University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Hong Kong Polytechnic University","institution_ids":["https://openalex.org/I14243506"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5129678183","display_name":"Yanda Zhu","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I881766915","display_name":"Nanjing University","ror":"https://ror.org/01rxvg760","country_code":"CN","type":"education","lineage":["https://openalex.org/I881766915"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Zhu, Yanda","raw_affiliation_strings":["Nanjing University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Nanjing University","institution_ids":["https://openalex.org/I881766915"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5021451616","display_name":"Yuanyang Zhu","orcid":"https://orcid.org/0000-0001-7484-1804","observed_orcids":["https://orcid.org/0000-0001-7484-1804"]},"institutions":[{"id":"https://openalex.org/I881766915","display_name":"Nanjing University","ror":"https://ror.org/01rxvg760","country_code":"CN","type":"education","lineage":["https://openalex.org/I881766915"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Zhu, Yuanyang","raw_affiliation_strings":["Nanjing University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Nanjing University","institution_ids":["https://openalex.org/I881766915"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5153528468","display_name":"Yaohua Hu","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I180726961","display_name":"Shenzhen University","ror":"https://ror.org/01vy4gh70","country_code":"CN","type":"education","lineage":["https://openalex.org/I180726961"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Hu, Yaohua","raw_affiliation_strings":["Shenzhen University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Shenzhen University","institution_ids":["https://openalex.org/I180726961"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5153542667","display_name":"Chunlin Chen","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I881766915","display_name":"Nanjing University","ror":"https://ror.org/01rxvg760","country_code":"CN","type":"education","lineage":["https://openalex.org/I881766915"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Chen, Chunlin","raw_affiliation_strings":["Nanjing University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Nanjing University","institution_ids":["https://openalex.org/I881766915"]}]}],"institutions":[],"countries_distinct_count":2,"institutions_distinct_count":3,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":true,"cited_by_count":0,"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":0.9980000257492065,"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":0.9980000257492065,"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/T12026","display_name":"Explainable Artificial Intelligence (XAI)","score":0.0017000000225380063,"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":9.999999747378752e-05,"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/multi-agent-reinforcement-learning","display_name":"multi-agent reinforcement learning","score":0.9480000138282776},{"id":"https://openalex.org/keywords/interpretability","display_name":"interpretability","score":0.9020000100135803},{"id":"https://openalex.org/keywords/latent-concepts","display_name":"latent concepts","score":0.902999997138977},{"id":"https://openalex.org/keywords/variational-autoencoder","display_name":"variational autoencoder","score":0.7850000262260437},{"id":"https://openalex.org/keywords/intrinsic-rewards","display_name":"intrinsic rewards","score":0.9409999847412109},{"id":"https://openalex.org/keywords/cooperative-multi-agent-systems","display_name":"cooperative multi-agent systems","score":0.5260000228881836},{"id":"https://openalex.org/keywords/partial-observability","display_name":"partial observability","score":0.7829999923706055}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7282999753952026},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.673799991607666},{"id":"https://openalex.org/C36299963","wikidata":"https://www.wikidata.org/wiki/Q1369844","display_name":"Observability","level":2,"score":0.6536999940872192},{"id":"https://openalex.org/C97541855","wikidata":"https://www.wikidata.org/wiki/Q830687","display_name":"Reinforcement learning","level":2,"score":0.642300009727478},{"id":"https://openalex.org/C108154423","wikidata":"https://www.wikidata.org/wiki/Q1469792","display_name":"Salience (neuroscience)","level":2,"score":0.6380000114440918},{"id":"https://openalex.org/C184898388","wikidata":"https://www.wikidata.org/wiki/Q1435712","display_name":"Pairwise comparison","level":2,"score":0.5920000076293945},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.48240000009536743},{"id":"https://openalex.org/C66746571","wikidata":"https://www.wikidata.org/wiki/Q1134833","display_name":"ENCODE","level":3,"score":0.48159998655319214},{"id":"https://openalex.org/C101738243","wikidata":"https://www.wikidata.org/wiki/Q786435","display_name":"Autoencoder","level":3,"score":0.45669999718666077},{"id":"https://openalex.org/C51167844","wikidata":"https://www.wikidata.org/wiki/Q4422623","display_name":"Latent variable","level":2,"score":0.41749998927116394},{"id":"https://openalex.org/C137002209","wikidata":"https://www.wikidata.org/wiki/Q898521","display_name":"Hidden variable theory","level":3,"score":0.382999986410141},{"id":"https://openalex.org/C2777212361","wikidata":"https://www.wikidata.org/wiki/Q5127848","display_name":"Class (philosophy)","level":2,"score":0.3449000120162964},{"id":"https://openalex.org/C2776359362","wikidata":"https://www.wikidata.org/wiki/Q2145286","display_name":"Representation (politics)","level":3,"score":0.33970001339912415},{"id":"https://openalex.org/C100776233","wikidata":"https://www.wikidata.org/wiki/Q2532492","display_name":"Bridge (graph theory)","level":2,"score":0.3208000063896179},{"id":"https://openalex.org/C65965080","wikidata":"https://www.wikidata.org/wiki/Q1806885","display_name":"Latent variable model","level":3,"score":0.2937999963760376},{"id":"https://openalex.org/C59404180","wikidata":"https://www.wikidata.org/wiki/Q17013334","display_name":"Feature learning","level":2,"score":0.29319998621940613},{"id":"https://openalex.org/C184337299","wikidata":"https://www.wikidata.org/wiki/Q1437428","display_name":"Semantics (computer science)","level":2,"score":0.29179999232292175},{"id":"https://openalex.org/C89611455","wikidata":"https://www.wikidata.org/wiki/Q6804646","display_name":"Mechanism (biology)","level":2,"score":0.2847000062465668},{"id":"https://openalex.org/C64543145","wikidata":"https://www.wikidata.org/wiki/Q162942","display_name":"Intersection (aeronautics)","level":2,"score":0.2831999957561493},{"id":"https://openalex.org/C13687954","wikidata":"https://www.wikidata.org/wiki/Q4826847","display_name":"Autonomous agent","level":2,"score":0.2782999873161316},{"id":"https://openalex.org/C113336015","wikidata":"https://www.wikidata.org/wiki/Q574010","display_name":"Complete information","level":2,"score":0.27720001339912415},{"id":"https://openalex.org/C2780695315","wikidata":"https://www.wikidata.org/wiki/Q3799040","display_name":"Unobservable","level":2,"score":0.2646999955177307}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2609.34459","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2609.34459","pdf_url":"https://arxiv.org/pdf/2609.34459","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.2609.34459","doi":"https://doi.org/10.48550/arxiv.2609.34459","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2609.34459","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":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"pmh:oai:arXiv.org:2609.34459","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2609.34459","pdf_url":"https://arxiv.org/pdf/2609.34459","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":[],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7214747136.pdf","grobid_xml":"https://content.openalex.org/works/W7214747136.grobid-xml"},"referenced_works_count":2,"referenced_works":["https://openalex.org/W3004640943","https://openalex.org/W4416168492"],"related_works":["https://openalex.org/W4408952435","https://openalex.org/W6979229092","https://openalex.org/W4321855176","https://openalex.org/W7065858195","https://openalex.org/W4407376604","https://openalex.org/W4417338844","https://openalex.org/W4327652961","https://openalex.org/W4417070065","https://openalex.org/W4385989051","https://openalex.org/W4225093764"],"abstract_inverted_index":{"Efficient":[0],"cooperation":[1],"is":[2],"challenging":[3],"due":[4],"to":[5,57,126,160],"the":[6,31,104,120,127,131],"usual":[7],"partial":[8],"observability":[9],"of":[10,106,122,163],"each":[11,63,107],"agent":[12,64],"in":[13,115,167],"multi-agent":[14],"reinforcement":[15],"learning.":[16],"Recurrent":[17],"networks":[18],"encode":[19],"local":[20,48,71,96],"interaction":[21],"histories,":[22],"but":[23,178],"their":[24,186],"hidden":[25],"representations":[26],"provide":[27],"limited":[28],"insight":[29],"into":[30,81],"information":[32],"underlying":[33],"individual":[34,123],"decisions.":[35,187],"To":[36,101],"address":[37],"these":[38],"challenges,":[39],"we":[40,109,142],"propose":[41],"a":[42,82,87,93,111,144],"novel":[43],"interpretable":[44],"framework,":[45],"called":[46],"escaping":[47],"views":[49,97],"(ELV),":[50],"which":[51,91,116,157],"introduces":[52],"semantically":[53],"structured":[54],"latent":[55,84],"concepts":[56,68,77,124],"render":[58],"policy":[59],"decisions":[60],"transparent.":[61],"Specifically,":[62],"extracts":[65],"low-dimensional":[66],"semantic":[67,164],"from":[69,153],"its":[70],"observation":[72],"and":[73,98],"action-observation":[74],"trajectory.":[75],"These":[76],"are":[78],"jointly":[79],"encoded":[80],"contextual":[83],"variable":[85],"via":[86],"variational":[88],"autoencoder":[89],"(VAE),":[90],"builds":[92],"bridge":[94],"between":[95],"global":[99,128],"semantics.":[100],"explicitly":[102,180],"model":[103],"decision":[105],"agent,":[108],"employ":[110],"dual-path":[112],"attention":[113],"mechanism":[114],"one":[117],"module":[118,147],"estimates":[119],"salience":[121],"relative":[125],"context,":[129],"while":[130],"other":[132],"captures":[133],"higher-order":[134],"cooperative":[135],"patterns":[136],"with":[137],"pairwise":[138],"concept":[139,145],"interactions.":[140],"Furthermore,":[141],"incorporate":[143],"prediction":[146,155],"that":[148,171],"derives":[149],"an":[150],"intrinsic":[151],"reward":[152],"next-concept":[154],"errors,":[156],"incentivizes":[158],"agents":[159,183],"explore":[161],"regions":[162],"novelty.":[165],"Experiments":[166],"multiple":[168],"environments":[169],"verify":[170],"ELV":[172],"not":[173],"only":[174],"achieves":[175],"competitive":[176],"performance":[177],"also":[179],"provides":[181],"how":[182],"reason":[184],"about":[185]},"counts_by_year":[],"updated_date":"2026-11-07T11:25:07.637180","created_date":"2026-09-30T00:00:00"}