{"id":"https://openalex.org/W7137860125","doi":"https://doi.org/10.1609/aaai.v40i38.40521","title":"Inference-Aware Prompt Optimization for Aligning Black-Box Large Language Models","display_name":"Inference-Aware Prompt Optimization for Aligning Black-Box Large Language Models","publication_year":2026,"publication_date":"2026-03-14","ids":{"openalex":"https://openalex.org/W7137860125","doi":"https://doi.org/10.1609/aaai.v40i38.40521"},"language":"en","primary_location":{"id":"doi:10.1609/aaai.v40i38.40521","is_oa":true,"landing_page_url":"https://doi.org/10.1609/aaai.v40i38.40521","pdf_url":"https://ojs.aaai.org/index.php/AAAI/article/download/40521/44482","source":{"id":"https://openalex.org/S4210191458","display_name":"Proceedings of the AAAI Conference on Artificial Intelligence","issn_l":"2159-5399","issn":["2159-5399","2374-3468"],"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/P4310320058","host_organization_name":"Association for the Advancement of Artificial Intelligence","host_organization_lineage":["https://openalex.org/P4310320058"],"host_organization_lineage_names":["Association for the Advancement of Artificial Intelligence"],"type":"conference"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the AAAI Conference on Artificial Intelligence","raw_type":"journal-article"},"type":"conference-paper","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"diamond","oa_url":"https://ojs.aaai.org/index.php/AAAI/article/download/40521/44482","any_repository_has_fulltext":false},"authorships":[{"author_position":"first","author":{"id":null,"display_name":"Saaduddin Mahmud","orcid":null},"institutions":[{"id":"https://openalex.org/I24603500","display_name":"University of Massachusetts Amherst","ror":"https://ror.org/0072zz521","country_code":"US","type":"education","lineage":["https://openalex.org/I24603500"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Saaduddin Mahmud","raw_affiliation_strings":["University of Massachusetts Amherst"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"University of Massachusetts Amherst","institution_ids":["https://openalex.org/I24603500"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Mason Nakamura","orcid":null},"institutions":[{"id":"https://openalex.org/I24603500","display_name":"University of Massachusetts Amherst","ror":"https://ror.org/0072zz521","country_code":"US","type":"education","lineage":["https://openalex.org/I24603500"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Mason Nakamura","raw_affiliation_strings":["University of Massachusetts Amherst"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"University of Massachusetts Amherst","institution_ids":["https://openalex.org/I24603500"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Kyle Hollins Wray","orcid":null},"institutions":[{"id":"https://openalex.org/I24603500","display_name":"University of Massachusetts Amherst","ror":"https://ror.org/0072zz521","country_code":"US","type":"education","lineage":["https://openalex.org/I24603500"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Kyle Hollins Wray","raw_affiliation_strings":["University of Massachusetts Amherst"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"University of Massachusetts Amherst","institution_ids":["https://openalex.org/I24603500"]}]},{"author_position":"last","author":{"id":null,"display_name":"Shlomo Zilberstein","orcid":null},"institutions":[{"id":"https://openalex.org/I24603500","display_name":"University of Massachusetts Amherst","ror":"https://ror.org/0072zz521","country_code":"US","type":"education","lineage":["https://openalex.org/I24603500"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Shlomo Zilberstein","raw_affiliation_strings":["University of Massachusetts Amherst"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"University of Massachusetts Amherst","institution_ids":["https://openalex.org/I24603500"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":["https://openalex.org/I24603500"],"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":"40","issue":"38","first_page":"32455","last_page":"32464"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10028","display_name":"Topic Modeling","score":0.4456000030040741,"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/T10028","display_name":"Topic Modeling","score":0.4456000030040741,"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/T10181","display_name":"Natural Language Processing Techniques","score":0.164000004529953,"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/T13629","display_name":"Text Readability and Simplification","score":0.05739999935030937,"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/inference","display_name":"Inference","score":0.8669999837875366},{"id":"https://openalex.org/keywords/task","display_name":"Task (project management)","score":0.4693000018596649},{"id":"https://openalex.org/keywords/voting","display_name":"Voting","score":0.3799000084400177},{"id":"https://openalex.org/keywords/computation","display_name":"Computation","score":0.36410000920295715},{"id":"https://openalex.org/keywords/sampling","display_name":"Sampling (signal processing)","score":0.35510000586509705},{"id":"https://openalex.org/keywords/language-model","display_name":"Language model","score":0.35420000553131104},{"id":"https://openalex.org/keywords/preference","display_name":"Preference","score":0.3294000029563904},{"id":"https://openalex.org/keywords/causal-inference","display_name":"Causal inference","score":0.3246999979019165}],"concepts":[{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.8669999837875366},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7861999869346619},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.5852000117301941},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.4984999895095825},{"id":"https://openalex.org/C2780451532","wikidata":"https://www.wikidata.org/wiki/Q759676","display_name":"Task (project management)","level":2,"score":0.4693000018596649},{"id":"https://openalex.org/C520049643","wikidata":"https://www.wikidata.org/wiki/Q189760","display_name":"Voting","level":3,"score":0.3799000084400177},{"id":"https://openalex.org/C45374587","wikidata":"https://www.wikidata.org/wiki/Q12525525","display_name":"Computation","level":2,"score":0.36410000920295715},{"id":"https://openalex.org/C140779682","wikidata":"https://www.wikidata.org/wiki/Q210868","display_name":"Sampling (signal processing)","level":3,"score":0.35510000586509705},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.35420000553131104},{"id":"https://openalex.org/C2781249084","wikidata":"https://www.wikidata.org/wiki/Q908656","display_name":"Preference","level":2,"score":0.3294000029563904},{"id":"https://openalex.org/C158600405","wikidata":"https://www.wikidata.org/wiki/Q5054566","display_name":"Causal inference","level":2,"score":0.3246999979019165},{"id":"https://openalex.org/C95167961","wikidata":"https://www.wikidata.org/wiki/Q4483495","display_name":"Fiducial inference","level":5,"score":0.3228999972343445},{"id":"https://openalex.org/C99844830","wikidata":"https://www.wikidata.org/wiki/Q102441924","display_name":"Scaling","level":2,"score":0.3222000002861023},{"id":"https://openalex.org/C120936955","wikidata":"https://www.wikidata.org/wiki/Q2155640","display_name":"Empirical research","level":2,"score":0.3003999888896942},{"id":"https://openalex.org/C137836250","wikidata":"https://www.wikidata.org/wiki/Q984063","display_name":"Optimization problem","level":2,"score":0.28369998931884766},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.2736000120639801},{"id":"https://openalex.org/C175154964","wikidata":"https://www.wikidata.org/wiki/Q380077","display_name":"Task analysis","level":3,"score":0.2721000015735626},{"id":"https://openalex.org/C166052673","wikidata":"https://www.wikidata.org/wiki/Q83021","display_name":"Empirical evidence","level":2,"score":0.2671000063419342},{"id":"https://openalex.org/C8505890","wikidata":"https://www.wikidata.org/wiki/Q605095","display_name":"Budget constraint","level":2,"score":0.2632000148296356},{"id":"https://openalex.org/C12725497","wikidata":"https://www.wikidata.org/wiki/Q810247","display_name":"Baseline (sea)","level":2,"score":0.2581000030040741},{"id":"https://openalex.org/C134261354","wikidata":"https://www.wikidata.org/wiki/Q938438","display_name":"Statistical inference","level":2,"score":0.25760000944137573}],"mesh":[],"locations_count":2,"locations":[{"id":"doi:10.1609/aaai.v40i38.40521","is_oa":true,"landing_page_url":"https://doi.org/10.1609/aaai.v40i38.40521","pdf_url":"https://ojs.aaai.org/index.php/AAAI/article/download/40521/44482","source":{"id":"https://openalex.org/S4210191458","display_name":"Proceedings of the AAAI Conference on Artificial Intelligence","issn_l":"2159-5399","issn":["2159-5399","2374-3468"],"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/P4310320058","host_organization_name":"Association for the Advancement of Artificial Intelligence","host_organization_lineage":["https://openalex.org/P4310320058"],"host_organization_lineage_names":["Association for the Advancement of Artificial Intelligence"],"type":"conference"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the AAAI Conference on Artificial Intelligence","raw_type":"journal-article"},{"id":"pmh:oai:ojs.aaai.org:article/40521","is_oa":false,"landing_page_url":"https://ojs.aaai.org/index.php/AAAI/article/view/40521","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"2159-5399","raw_type":"info:eu-repo/semantics/article"}],"best_oa_location":{"id":"doi:10.1609/aaai.v40i38.40521","is_oa":true,"landing_page_url":"https://doi.org/10.1609/aaai.v40i38.40521","pdf_url":"https://ojs.aaai.org/index.php/AAAI/article/download/40521/44482","source":{"id":"https://openalex.org/S4210191458","display_name":"Proceedings of the AAAI Conference on Artificial Intelligence","issn_l":"2159-5399","issn":["2159-5399","2374-3468"],"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/P4310320058","host_organization_name":"Association for the Advancement of Artificial Intelligence","host_organization_lineage":["https://openalex.org/P4310320058"],"host_organization_lineage_names":["Association for the Advancement of Artificial Intelligence"],"type":"conference"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the AAAI Conference on Artificial Intelligence","raw_type":"journal-article"},"sustainable_development_goals":[{"id":"https://metadata.un.org/sdg/17","display_name":"Partnerships for the goals","score":0.4932979643344879}],"awards":[{"id":"https://openalex.org/G3637287070","display_name":"RI: Small: Foundations and Applications of Observer-Aware Planning","funder_award_id":"2205153","funder_id":"https://openalex.org/F4320306076","funder_display_name":"National Science Foundation"},{"id":"https://openalex.org/G6534177744","display_name":"Collaborative Research: SLES: No Bad Surprises: Aligning Agent and Human Norms via Specification Refinements","funder_award_id":"2416460","funder_id":"https://openalex.org/F4320306076","funder_display_name":"National Science Foundation"},{"id":"https://openalex.org/G871475615","display_name":"CISE: Large: Causal Foundations for Decision Making and Learning","funder_award_id":"2321786","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/F4320337807","display_name":"U.S. Army Combat Capabilities Development Command","ror":null}],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7137860125.pdf","grobid_xml":"https://content.openalex.org/works/W7137860125.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Prompt":[0,118],"optimization":[1,45],"methods":[2],"have":[3,26],"demonstrated":[4],"significant":[5,65],"effectiveness":[6,167],"in":[7,187],"aligning":[8,188],"black-box":[9,189],"large":[10],"language":[11],"models":[12],"(LLMs).":[13],"In":[14],"parallel,":[15],"inference":[16,48,60,94,103,126,133],"scaling":[17],"strategies":[18],"such":[19],"as":[20,68],"Best-of-N":[21],"Sampling":[22],"and":[23,33,71,93,102,125,135,155,177,179],"Majority":[24],"Voting":[25],"likewise":[27],"been":[28],"shown":[29],"to":[30],"improve":[31],"alignment":[32],"performance":[34],"by":[35],"trading":[36],"additional":[37],"computation":[38],"for":[39,58,146],"better":[40],"output.":[41],"However,":[42],"existing":[43],"prompt":[44,101,124,192],"approaches":[46],"are":[47],"strategy":[49],"agnostic;":[50],"that":[51,85,120],"is,":[52],"they":[53],"optimize":[54],"prompts":[55],"without":[56],"accounting":[57],"the":[59,98,123,132,160,166,181],"strategy.":[61],"This":[62],"constitutes":[63],"a":[64,75,111,142],"methodological":[66],"gap,":[67,108],"our":[69],"empirical":[70],"theoretical":[72],"analysis":[73],"reveals":[74],"strong":[76],"interdependence":[77],"between":[78],"these":[79],"two":[80],"paradigms.":[81],"Moreover,":[82],"we":[83,109,164],"find":[84],"user":[86],"preferences":[87],"regarding":[88],"trade-offs":[89],"among":[90],"multiple":[91],"objectives":[92],"budgets":[95],"substantially":[96],"influence":[97],"choice":[99],"of":[100,131,168,184],"configuration.":[104],"To":[105],"address":[106],"this":[107],"introduce":[110],"novel":[112],"unified":[113],"framework":[114],"named":[115],"IAPO":[116],"(Inference-Aware":[117],"Optimization)":[119],"jointly":[121],"optimizes":[122],"scale,":[127],"while":[128],"being":[129],"aware":[130],"budget":[134],"different":[136],"task":[137],"objectives.":[138],"We":[139],"then":[140],"develop":[141],"fixed-budget":[143],"training":[144],"algorithm":[145],"IAPO,":[147],"called":[148],"PSST":[149,169],"(Prompt":[150],"Scaling":[151],"via":[152],"Sequential":[153],"Trimming),":[154],"establish":[156],"finite-budget":[157],"guarantees":[158],"on":[159,170],"error":[161],"probability.":[162],"Finally,":[163],"evaluate":[165],"six":[171],"tasks,":[172],"including":[173],"multi-objective":[174],"text":[175],"generation":[176],"reasoning,":[178],"demonstrate":[180],"critical":[182],"role":[183],"incorporating":[185],"inference-awareness":[186],"LLMs":[190],"using":[191],"optimization.":[193]},"counts_by_year":[],"updated_date":"2026-07-14T23:27:15.235271","created_date":"2026-02-06T00:00:00"}