{"id":"https://openalex.org/W7166843738","doi":"https://doi.org/10.18653/v1/2026.findings-acl.1858","title":"SemPA: Improving Sentence Embeddings of Large Language Models through Semantic Preference Alignment","display_name":"SemPA: Improving Sentence Embeddings of Large Language Models through Semantic Preference Alignment","publication_year":2026,"publication_date":"2026-01-01","ids":{"openalex":"https://openalex.org/W7166843738","doi":"https://doi.org/10.18653/v1/2026.findings-acl.1858"},"language":null,"primary_location":{"id":"doi:10.18653/v1/2026.findings-acl.1858","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.findings-acl.1858","pdf_url":"https://aclanthology.org/2026.findings-acl.1858.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: ACL 2026","raw_type":"proceedings-article"},"type":"conference-paper","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://aclanthology.org/2026.findings-acl.1858.pdf","any_repository_has_fulltext":null},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5139738629","display_name":"Ziyang Chen","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ziyang Chen","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5122450140","display_name":"Zhenxuan Huang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhenxuan Huang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139790574","display_name":"Yile Wang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yile Wang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139829070","display_name":"Weiqin Wang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Weiqin Wang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139750562","display_name":"Lu Yin","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Lu Yin","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5139824359","display_name":"Hui Huang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Hui Huang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":0,"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":0.83378551,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"37299","last_page":"37312"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10664","display_name":"Sentiment Analysis and Opinion Mining","score":0.41589999198913574,"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/T10664","display_name":"Sentiment Analysis and Opinion Mining","score":0.41589999198913574,"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":0.2506999969482422,"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":0.028200000524520874,"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/sentence","display_name":"Sentence","score":0.5655999779701233},{"id":"https://openalex.org/keywords/preference","display_name":"Preference","score":0.47749999165534973},{"id":"https://openalex.org/keywords/semantics","display_name":"Semantics (computer science)","score":0.3939000070095062},{"id":"https://openalex.org/keywords/language-model","display_name":"Language model","score":0.357699990272522},{"id":"https://openalex.org/keywords/natural-language","display_name":"Natural language","score":0.3407999873161316},{"id":"https://openalex.org/keywords/feature","display_name":"Feature (linguistics)","score":0.3393999934196472}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6743999719619751},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.6301000118255615},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.6051999926567078},{"id":"https://openalex.org/C2777530160","wikidata":"https://www.wikidata.org/wiki/Q41796","display_name":"Sentence","level":2,"score":0.5655999779701233},{"id":"https://openalex.org/C2781249084","wikidata":"https://www.wikidata.org/wiki/Q908656","display_name":"Preference","level":2,"score":0.47749999165534973},{"id":"https://openalex.org/C184337299","wikidata":"https://www.wikidata.org/wiki/Q1437428","display_name":"Semantics (computer science)","level":2,"score":0.3939000070095062},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.357699990272522},{"id":"https://openalex.org/C195324797","wikidata":"https://www.wikidata.org/wiki/Q33742","display_name":"Natural language","level":2,"score":0.3407999873161316},{"id":"https://openalex.org/C2776401178","wikidata":"https://www.wikidata.org/wiki/Q12050496","display_name":"Feature (linguistics)","level":2,"score":0.3393999934196472},{"id":"https://openalex.org/C41895202","wikidata":"https://www.wikidata.org/wiki/Q8162","display_name":"Linguistics","level":1,"score":0.3140999972820282},{"id":"https://openalex.org/C124304363","wikidata":"https://www.wikidata.org/wiki/Q673661","display_name":"Abstraction","level":2,"score":0.2946000099182129},{"id":"https://openalex.org/C155092808","wikidata":"https://www.wikidata.org/wiki/Q182557","display_name":"Computational linguistics","level":2,"score":0.2531999945640564},{"id":"https://openalex.org/C2983448237","wikidata":"https://www.wikidata.org/wiki/Q1078276","display_name":"Language understanding","level":2,"score":0.25040000677108765}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.18653/v1/2026.findings-acl.1858","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.findings-acl.1858","pdf_url":"https://aclanthology.org/2026.findings-acl.1858.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: ACL 2026","raw_type":"proceedings-article"}],"best_oa_location":{"id":"doi:10.18653/v1/2026.findings-acl.1858","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.findings-acl.1858","pdf_url":"https://aclanthology.org/2026.findings-acl.1858.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: ACL 2026","raw_type":"proceedings-article"},"sustainable_development_goals":[],"awards":[],"funders":[{"id":"https://openalex.org/F4320316880","display_name":"Natural Science Foundation of Shenzhen City","ror":null},{"id":"https://openalex.org/F4320325571","display_name":"Shenzhen University","ror":"https://ror.org/01vy4gh70"},{"id":"https://openalex.org/F4320337111","display_name":"Basic and Applied Basic Research Foundation of Guangdong Province","ror":null}],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7166843738.pdf","grobid_xml":"https://content.openalex.org/works/W7166843738.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Traditional":[0],"sentence":[1,74],"embedding":[2,15],"methods":[3,16,24],"employ":[4],"token-level":[5],"contrastive":[6,124],"learning":[7,125],"on":[8,18,27,97,133],"nongenerative":[9],"pre-trained":[10],"models.Recently,":[11],"there":[12],"have":[13],"emerged":[14],"based":[17],"generative":[19,64,79,114],"large":[20],"language":[21],"models":[22],"(LLMs).These":[23],"either":[25],"rely":[26],"fixed":[28],"prompt":[29],"templates":[30],"or":[31],"involve":[32],"modifications":[33],"to":[34,93,106],"the":[35,43,51,54,59,73,78,103,127,153,160],"model":[36,44,104,129],"architecture.The":[37],"former":[38],"lacks":[39],"further":[40],"optimization":[41],"of":[42,58,81,157],"and":[45,123,139],"results":[46,132],"in":[47],"limited":[48],"performance,":[49],"while":[50,76,111],"latter":[52],"alters":[53],"internal":[55],"computational":[56],"mechanisms":[57],"model,":[60],"thereby":[61],"compromising":[62],"its":[63],"capabilities.We":[65],"propose":[66],"SEMPA,":[67],"a":[68,98,118],"novel":[69],"approach":[70],"that":[71,145],"boosts":[72],"representations":[75,150],"preserving":[77,112],"ability":[80],"LLMs":[82,96,143],"via":[83],"semantic":[84,135,149],"preference":[85],"alignment.We":[86],"leverage":[87],"sentence-level":[88],"Direct":[89],"Preference":[90],"Optimization":[91],"(DPO)":[92],"efficiently":[94],"optimize":[95],"paraphrase":[99],"generation":[100,155],"task,":[101],"where":[102],"learns":[105],"discriminate":[107],"semantically":[108],"equivalent":[109],"sentences":[110],"inherent":[113,154],"capacity.Theoretically,":[115],"we":[116],"establish":[117],"formal":[119],"connection":[120],"between":[121],"DPO":[122],"under":[126],"Plackett-Luce":[128],"framework.Empirically,":[130],"experimental":[131],"both":[134],"textual":[136],"similarity":[137],"tasks":[138],"various":[140],"benchmarks":[141],"for":[142],"show":[144],"SEMPA":[146],"achieves":[147],"better":[148],"without":[151],"sacrificing":[152],"capability":[156],"LLMs.We":[158],"release":[159],"code":[161],"at":[162],"https:":[163],"//github.com/szu-tera/SemPA.":[164]},"counts_by_year":[],"updated_date":"2026-07-29T14:22:42.915294","created_date":"2026-07-02T00:00:00"}