{"id":"https://openalex.org/W7140094036","doi":"https://doi.org/10.18653/v1/2026.findings-eacl.233","title":"MAPRO: Recasting Multi-Agent Prompt Optimization as Maximum a Posteriori Inference","display_name":"MAPRO: Recasting Multi-Agent Prompt Optimization as Maximum a Posteriori Inference","publication_year":2026,"publication_date":"2026-01-01","ids":{"openalex":"https://openalex.org/W7140094036","doi":"https://doi.org/10.18653/v1/2026.findings-eacl.233"},"language":"en","primary_location":{"id":"doi:10.18653/v1/2026.findings-eacl.233","doi":"https://doi.org/10.18653/v1/2026.findings-eacl.233","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.findings-eacl.233","pdf_url":"https://aclanthology.org/2026.findings-eacl.233.pdf","source":{"id":"https://openalex.org/S7407087051","display_name":"Findings of the Association for Computational Linguistics: EACL","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":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Findings of the Association for Computational Linguistics: EACL 2026","raw_type":"proceedings-article"},"volume":null,"type":"conference-paper","indexed_in":["arxiv","crossref","datacite"],"open_access":{"is_oa":true,"oa_status":"hybrid","oa_url":"https://aclanthology.org/2026.findings-eacl.233.pdf","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5130365323","display_name":"Zheyuan Zhang","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I107639228","display_name":"University of Notre Dame","ror":"https://ror.org/00mkhxb43","country_code":"US","type":"education","lineage":["https://openalex.org/I107639228"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Zheyuan Zhang","raw_affiliation_strings":["University of Notre Dame ,"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"University of Notre Dame ,","institution_ids":["https://openalex.org/I107639228"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5125636658","display_name":"Lin Ge","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I1311688040","display_name":"Amazon (United States)","ror":"https://ror.org/04mv4n011","country_code":"US","type":"company","lineage":["https://openalex.org/I1311688040"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Lin Ge","raw_affiliation_strings":["Amazon"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Amazon","institution_ids":["https://openalex.org/I1311688040"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5130378634","display_name":"Hongjiang Li","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I1311688040","display_name":"Amazon (United States)","ror":"https://ror.org/04mv4n011","country_code":"US","type":"company","lineage":["https://openalex.org/I1311688040"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Hongjiang Li","raw_affiliation_strings":["Amazon"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Amazon","institution_ids":["https://openalex.org/I1311688040"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5112488574","display_name":"Weicheng Zhu","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I1311688040","display_name":"Amazon (United States)","ror":"https://ror.org/04mv4n011","country_code":"US","type":"company","lineage":["https://openalex.org/I1311688040"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Weicheng Zhu","raw_affiliation_strings":["Amazon"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Amazon","institution_ids":["https://openalex.org/I1311688040"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5022275632","display_name":"Chuxu Zhang","orcid":"https://orcid.org/0000-0002-8349-7926","observed_orcids":["https://orcid.org/0000-0002-8349-7926"]},"institutions":[{"id":"https://openalex.org/I140172145","display_name":"University of Connecticut","ror":"https://ror.org/02der9h97","country_code":"US","type":"education","lineage":["https://openalex.org/I140172145"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Chuxu Zhang","raw_affiliation_strings":["University of Connecticut ,"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"University of Connecticut ,","institution_ids":["https://openalex.org/I140172145"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5130388966","display_name":"Yanfang Ye","orcid":"https://orcid.org/0009-0003-6002-125X","observed_orcids":["https://orcid.org/0009-0003-6002-125X"]},"institutions":[{"id":"https://openalex.org/I107639228","display_name":"University of Notre Dame","ror":"https://ror.org/00mkhxb43","country_code":"US","type":"education","lineage":["https://openalex.org/I107639228"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Yanfang Ye","raw_affiliation_strings":["University of Notre Dame ,"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"University of Notre Dame ,","institution_ids":["https://openalex.org/I107639228"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":3,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":17.0938,"has_fulltext":true,"cited_by_count":2,"citation_normalized_percentile":{"value":0.98780943,"is_in_top_1_percent":false,"is_in_top_10_percent":true},"cited_by_percentile_year":{"min":96,"max":97},"biblio":{"volume":null,"issue":null,"first_page":"4458","last_page":"4480"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10181","display_name":"Natural Language Processing Techniques","score":0.37450000643730164,"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/T10181","display_name":"Natural Language Processing Techniques","score":0.37450000643730164,"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/T10456","display_name":"Multi-Agent Systems and Negotiation","score":0.23409999907016754,"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/T10028","display_name":"Topic Modeling","score":0.18150000274181366,"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-systems","display_name":"multi-agent systems","score":0.8830000162124634},{"id":"https://openalex.org/keywords/prompt-optimization","display_name":"prompt optimization","score":0.9290000200271606},{"id":"https://openalex.org/keywords/maximum-a-posteriori-estimation","display_name":"maximum a posteriori estimation","score":0.9010000228881836},{"id":"https://openalex.org/keywords/credit-assignment","display_name":"credit assignment","score":0.6629999876022339},{"id":"https://openalex.org/keywords/llm-based-agents","display_name":"LLM agents","score":0.6840000152587891},{"id":"https://openalex.org/keywords/prompt-sensitivity","display_name":"prompt sensitivity","score":0.527999997138977},{"id":"https://openalex.org/keywords/execution-feedback","display_name":"execution feedback","score":0.578000009059906}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.5264999866485596},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.460999995470047},{"id":"https://openalex.org/C75553542","wikidata":"https://www.wikidata.org/wiki/Q178161","display_name":"A priori and a posteriori","level":2,"score":0.45260000228881836},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.40130001306533813},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.37560001015663147},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.3434999883174896},{"id":"https://openalex.org/C126255220","wikidata":"https://www.wikidata.org/wiki/Q141495","display_name":"Mathematical optimization","level":1,"score":0.3208000063896179},{"id":"https://openalex.org/C9810830","wikidata":"https://www.wikidata.org/wiki/Q635384","display_name":"Maximum a posteriori estimation","level":3,"score":0.30090001225471497},{"id":"https://openalex.org/C137836250","wikidata":"https://www.wikidata.org/wiki/Q984063","display_name":"Optimization problem","level":2,"score":0.29409998655319214},{"id":"https://openalex.org/C147764199","wikidata":"https://www.wikidata.org/wiki/Q6865248","display_name":"Minification","level":2,"score":0.26460000872612},{"id":"https://openalex.org/C2776401178","wikidata":"https://www.wikidata.org/wiki/Q12050496","display_name":"Feature (linguistics)","level":2,"score":0.251800000667572}],"mesh":[],"locations_count":3,"locations":[{"id":"doi:10.18653/v1/2026.findings-eacl.233","doi":"https://doi.org/10.18653/v1/2026.findings-eacl.233","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.findings-eacl.233","pdf_url":"https://aclanthology.org/2026.findings-eacl.233.pdf","source":{"id":"https://openalex.org/S7407087051","display_name":"Findings of the Association for Computational Linguistics: EACL","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":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Findings of the Association for Computational Linguistics: EACL 2026","raw_type":"proceedings-article"},{"id":"pmh:oai:arXiv.org:2510.07475","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2510.07475","pdf_url":"https://arxiv.org/pdf/2510.07475","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.2510.07475","doi":"https://doi.org/10.48550/arxiv.2510.07475","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2510.07475","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":"doi:10.18653/v1/2026.findings-eacl.233","doi":"https://doi.org/10.18653/v1/2026.findings-eacl.233","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.findings-eacl.233","pdf_url":"https://aclanthology.org/2026.findings-eacl.233.pdf","source":{"id":"https://openalex.org/S7407087051","display_name":"Findings of the Association for Computational Linguistics: EACL","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":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Findings of the Association for Computational Linguistics: EACL 2026","raw_type":"proceedings-article"},"sustainable_development_goals":[],"sustainable_development_goals_aurora":[],"study_designs":[],"awards":[{"id":"https://openalex.org/G1601924855","display_name":"III: Small: A New Machine Learning Paradigm Towards Effective yet Efficient Foundation Graph Learning Models","funder_award_id":"2321504","funder_id":"https://openalex.org/F4320306076","funder_display_name":"National Science Foundation"},{"id":"https://openalex.org/G2320213723","display_name":"SCC-IRG Track 1: A New AI-driven Paradigm to Promote Community Resilience for Teenagers and Young Adults in Preventing Opioid Misuse and Addiction","funder_award_id":"2426514","funder_id":"https://openalex.org/F4320306076","funder_display_name":"National Science Foundation"},{"id":"https://openalex.org/G2325374520","display_name":"Collaborative Research: III: Small: Data-driven and AI-powered Innovations for Tackling the Nexus of Poverty-driven Food Insecurity and the Opioid Crisis","funder_award_id":"2533550","funder_id":"https://openalex.org/F4320306076","funder_display_name":"National Science Foundation"},{"id":"https://openalex.org/G3087910168","display_name":"III: Medium: A Data-driven and AI-augmented Framework for Collaborative Decision Making to Combat Infectious Disease Outbreaks","funder_award_id":"2217239","funder_id":"https://openalex.org/F4320306076","funder_display_name":"National Science Foundation"},{"id":"https://openalex.org/G6745018468","display_name":"D-ISN: An AI-augmented Framework to Detect, Disrupt, and Dismantle Opioid Trafficking Networks","funder_award_id":"2146076","funder_id":"https://openalex.org/F4320306076","funder_display_name":"National Science Foundation"}],"funders":[{"id":"https://openalex.org/F4320306076","display_name":"National Science Foundation","ror":"https://ror.org/021nxhr62"},{"id":"https://openalex.org/F4320337391","display_name":"Division of Civil, Mechanical and Manufacturing Innovation","ror":"https://ror.org/028yd4c30"}],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7140094036.pdf","grobid_xml":"https://content.openalex.org/works/W7140094036.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":["https://openalex.org/W7160665229","https://openalex.org/W7165628733","https://openalex.org/W7212009224","https://openalex.org/W4407186876","https://openalex.org/W7133531098","https://openalex.org/W7162797406","https://openalex.org/W7213967155","https://openalex.org/W7166526419","https://openalex.org/W7138533873","https://openalex.org/W7161154098"],"abstract_inverted_index":{"Large":[0],"language":[1],"models":[2],"(LLMs)":[3],"have":[4,62],"demonstrated":[5],"remarkable":[6],"capabilities":[7],"across":[8],"diverse":[9],"tasks,":[10,166],"and":[11,46,77,112,127,142,176,192],"LLM-based":[12],"agents":[13,32],"further":[14],"extend":[15],"these":[16],"abilities":[17],"to":[18,43,145,155],"various":[19,165],"practical":[20],"workflows.While":[21],"recent":[22,56,177],"progress":[23],"shows":[24],"that":[25,98,138],"multi-agent":[26,66,194],"systems":[27,195],"(MAS)":[28],"can":[29],"outperform":[30],"single":[31],"by":[33],"coordinating":[34],"specialized":[35],"roles,":[36],"designing":[37],"effective":[38],"MAS":[39,50,101],"remains":[40,69],"difficult":[41],"due":[42],"prompt":[44,60,67,102,161],"sensitivity":[45],"the":[47,54,129,197],"compounded":[48],"instability":[49],"creates.To":[51],"cope":[52],"with":[53],"challenge,":[55],"efforts":[57],"in":[58,164,196],"automated":[59,178],"design":[61,84],"reduced":[63],"manual":[64],"effort.However,":[65],"optimization":[68,103],"largely":[70],"unexplored.Challenges":[71],"like":[72],"exponentially":[73],"expanding":[74],"search":[75],"space":[76],"ambiguous":[78],"credit":[79,125],"assignment":[80,126],"together":[81],"make":[82],"systematic":[83],"intractable":[85],"without":[86],"principled":[87,193],"methods.Therefore,":[88],"we":[89],"introduce":[90],"Multi-Agent":[91],"PRompt":[92],"Optimization":[93],"(MAPRO),":[94],"a":[95,105,107,116,134,156],"four-stage":[96],"framework":[97],"first":[99],"formulates":[100],"as":[104],"Maximum":[106],"Posteriori":[108],"(MAP)":[109],"inference":[110],"problem":[111],"solves":[113],"it":[114],"using":[115],"languageguided":[117],"variant":[118],"of":[119,159],"max-product":[120],"belief":[121],"propagation":[122],"algorithm.To":[123],"address":[124],"updates":[128],"system":[130],"iteratively,":[131],"MAPRO":[132,152,167],"employs":[133],"topology-aware":[135],"refinement":[136],"mechanism":[137],"integrates":[139],"execution":[140],"feedback":[141],"downstream":[143],"blames":[144],"selectively":[146],"update":[147],"agent":[148],"prompts.Through":[149],"this":[150],"process,":[151],"progressively":[153],"converges":[154],"coordinated":[157],"set":[158],"agent-specific":[160],"policies.Across":[162],"benchmarks":[163],"achieves":[168],"stateof-the-art":[169],"performance,":[170,180],"consistently":[171],"surpassing":[172],"manually":[173],"engineered":[174],"baselines":[175],"alternatives.Beyond":[179],"our":[181],"MAP-based":[182],"formulation":[183],"also":[184],"delivers":[185],"general":[186],"guidelines":[187],"for":[188],"building":[189],"more":[190],"reliable":[191],"future":[198],"1":[199],".":[200]},"counts_by_year":[{"year":2026,"cited_by_count":2}],"updated_date":"2026-10-12T09:40:04.860316","created_date":"2026-03-24T00:00:00"}