{"id":"https://openalex.org/W4416034625","doi":"https://doi.org/10.18653/v1/2025.findings-emnlp.795","title":"PMPO: Probabilistic Metric Prompt Optimization for Small and Large Language Models","display_name":"PMPO: Probabilistic Metric Prompt Optimization for Small and Large Language Models","publication_year":2025,"publication_date":"2025-01-01","ids":{"openalex":"https://openalex.org/W4416034625","doi":"https://doi.org/10.18653/v1/2025.findings-emnlp.795"},"language":null,"primary_location":{"id":"doi:10.18653/v1/2025.findings-emnlp.795","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.findings-emnlp.795","pdf_url":"https://aclanthology.org/2025.findings-emnlp.795.pdf","source":null,"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: EMNLP 2025","raw_type":"proceedings-article"},"type":"conference-paper","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://aclanthology.org/2025.findings-emnlp.795.pdf","any_repository_has_fulltext":null},"authorships":[{"author_position":"first","author":{"id":null,"display_name":"ChenZhuo Zhao","orcid":null},"institutions":[{"id":"https://openalex.org/I20231570","display_name":"Peking University","ror":"https://ror.org/02v51f717","country_code":"CN","type":"education","lineage":["https://openalex.org/I20231570"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"ChenZhuo Zhao","raw_affiliation_strings":["Peking University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Peking University","institution_ids":["https://openalex.org/I20231570"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5087235476","display_name":"Ziqian Liu","orcid":"https://orcid.org/0000-0002-2041-7159"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ziqian Liu","raw_affiliation_strings":["Independent Researcher"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Independent Researcher","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5008458015","display_name":"Xinda Wang","orcid":"https://orcid.org/0000-0002-2735-2490"},"institutions":[{"id":"https://openalex.org/I20231570","display_name":"Peking University","ror":"https://ror.org/02v51f717","country_code":"CN","type":"education","lineage":["https://openalex.org/I20231570"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Xinda Wang","raw_affiliation_strings":["Peking University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Peking University","institution_ids":["https://openalex.org/I20231570"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5033874355","display_name":"Junting Lu","orcid":"https://orcid.org/0000-0001-7858-6679"},"institutions":[{"id":"https://openalex.org/I20231570","display_name":"Peking University","ror":"https://ror.org/02v51f717","country_code":"CN","type":"education","lineage":["https://openalex.org/I20231570"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Junting Lu","raw_affiliation_strings":["Peking University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Peking University","institution_ids":["https://openalex.org/I20231570"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5120308642","display_name":"Chaoyi Ruan","orcid":null},"institutions":[{"id":"https://openalex.org/I165932596","display_name":"National University of Singapore","ror":"https://ror.org/01tgyzw49","country_code":"SG","type":"education","lineage":["https://openalex.org/I165932596"]}],"countries":["SG"],"is_corresponding":false,"raw_author_name":"Chaoyi Ruan","raw_affiliation_strings":["National University of Singapore"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"National University of Singapore","institution_ids":["https://openalex.org/I165932596"]}]}],"institutions":[],"countries_distinct_count":2,"institutions_distinct_count":2,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":true,"cited_by_count":0,"citation_normalized_percentile":{"value":0.29050934,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"14728","last_page":"14761"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10181","display_name":"Natural Language Processing Techniques","score":0.3215000033378601,"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":0.3215000033378601,"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":0.1664000004529953,"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/T10201","display_name":"Speech Recognition and Synthesis","score":0.05169999971985817,"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/probabilistic-logic","display_name":"Probabilistic logic","score":0.5584999918937683},{"id":"https://openalex.org/keywords/metric","display_name":"Metric (unit)","score":0.5350000262260437},{"id":"https://openalex.org/keywords/statistical-model","display_name":"Statistical model","score":0.314300000667572},{"id":"https://openalex.org/keywords/probabilistic-relevance-model","display_name":"Probabilistic relevance model","score":0.29829999804496765},{"id":"https://openalex.org/keywords/constraint","display_name":"Constraint (computer-aided design)","score":0.2793999910354614}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6082000136375427},{"id":"https://openalex.org/C49937458","wikidata":"https://www.wikidata.org/wiki/Q2599292","display_name":"Probabilistic logic","level":2,"score":0.5584999918937683},{"id":"https://openalex.org/C176217482","wikidata":"https://www.wikidata.org/wiki/Q860554","display_name":"Metric (unit)","level":2,"score":0.5350000262260437},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.4023999869823456},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.3790999948978424},{"id":"https://openalex.org/C114289077","wikidata":"https://www.wikidata.org/wiki/Q3284399","display_name":"Statistical model","level":2,"score":0.314300000667572},{"id":"https://openalex.org/C143017306","wikidata":"https://www.wikidata.org/wiki/Q3318133","display_name":"Probabilistic relevance model","level":4,"score":0.29829999804496765},{"id":"https://openalex.org/C2776036281","wikidata":"https://www.wikidata.org/wiki/Q48769818","display_name":"Constraint (computer-aided design)","level":2,"score":0.2793999910354614},{"id":"https://openalex.org/C2778334786","wikidata":"https://www.wikidata.org/wiki/Q1586270","display_name":"Variation (astronomy)","level":2,"score":0.2727000117301941},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.2680000066757202}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.18653/v1/2025.findings-emnlp.795","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.findings-emnlp.795","pdf_url":"https://aclanthology.org/2025.findings-emnlp.795.pdf","source":null,"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: EMNLP 2025","raw_type":"proceedings-article"}],"best_oa_location":{"id":"doi:10.18653/v1/2025.findings-emnlp.795","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.findings-emnlp.795","pdf_url":"https://aclanthology.org/2025.findings-emnlp.795.pdf","source":null,"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: EMNLP 2025","raw_type":"proceedings-article"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4416034625.pdf","grobid_xml":"https://content.openalex.org/works/W4416034625.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Prompt":[0,55],"optimization":[1],"is":[2],"a":[3,57,67,78,100],"practical":[4],"and":[5,82,107,131,137,156,159,174],"widely":[6],"applicable":[7],"alternative":[8],"to":[9,86,121],"fine":[10],"tuning":[11],"for":[12,41,113],"improving":[13],"large":[14],"language":[15],"model":[16,135],"performance.Yet":[17],"many":[18],"existing":[19],"methods":[20],"evaluate":[21],"candidate":[22],"prompts":[23],"by":[24,96,165],"sampling":[25,106],"full":[26],"outputs,":[27],"often":[28],"coupled":[29],"with":[30],"self":[31],"critique":[32],"or":[33,44,109],"human":[34,108],"annotated":[35],"preferences,":[36],"which":[37],"limits":[38],"scalability,":[39],"especially":[40],"smaller":[42],"models":[43,45],"that":[46,60],"are":[47],"not":[48],"instruction":[49],"tuned.We":[50],"present":[51],"PMPO":[52,92,139],"(Probabilistic":[53],"Metric":[54],"Optimization),":[56],"unified":[58],"framework":[59],"uses":[61],"token":[62],"level":[63],"cross":[64],"entropy":[65],"as":[66],"direct,":[68],"lightweight":[69],"evaluation":[70],"signal.PMPO":[71],"locates":[72],"low":[73],"quality":[74],"prompt":[75,142],"segments":[76],"via":[77],"masking":[79],"based":[80,111,126,133],"analysis":[81],"iteratively":[83],"rewrites":[84],"them":[85],"propose":[87,122],"improved":[88],"variants.Crucially,":[89],"during":[90],"evaluation,":[91],"selects":[93],"among":[94],"variants":[95],"minimizing":[97],"loss":[98,125],"in":[99],"single":[101],"forward":[102],"pass,":[103],"eliminating":[104],"output":[105],"judge":[110],"scoring":[112],"selection":[114],"while":[115],"still":[116],"using":[117],"standard":[118],"generation":[119],"only":[120],"rewrites.This":[123],"unified,":[124],"strategy":[127],"supports":[128],"both":[129],"supervised":[130],"preference":[132],"tasks.Across":[134],"sizes":[136],"datasets,":[138],"outperforms":[140],"prior":[141],"optimizers:":[143],"it":[144],"achieves":[145],"the":[146],"highest":[147],"average":[148],"accuracy":[149],"on":[150,154],"BBH,":[151],"performs":[152],"strongly":[153],"GSM8K":[155],"AQuA":[157],"RAT,":[158],"raises":[160],"Al-pacaEval":[161],"2.0":[162],"win":[163],"rates":[164],"over":[166],"19":[167],"points.These":[168],"results":[169],"demonstrate":[170],"PMPO's":[171],"effectiveness,":[172],"efficiency,":[173],"broad":[175],"applicability.":[176]},"counts_by_year":[],"updated_date":"2026-08-22T07:34:49.880490","created_date":"2025-11-08T00:00:00"}