{"id":"https://openalex.org/W4416036849","doi":"https://doi.org/10.18653/v1/2025.emnlp-main.213","title":"Diffusion vs. Autoregressive Language Models: A Text Embedding Perspective","display_name":"Diffusion vs. Autoregressive Language Models: A Text Embedding Perspective","publication_year":2025,"publication_date":"2025-01-01","ids":{"openalex":"https://openalex.org/W4416036849","doi":"https://doi.org/10.18653/v1/2025.emnlp-main.213"},"language":null,"primary_location":{"id":"doi:10.18653/v1/2025.emnlp-main.213","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.emnlp-main.213","pdf_url":"https://aclanthology.org/2025.emnlp-main.213.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":"Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing","raw_type":"proceedings-article"},"type":"article","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://aclanthology.org/2025.emnlp-main.213.pdf","any_repository_has_fulltext":null},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5101897994","display_name":"Siyue Zhang","orcid":"https://orcid.org/0000-0002-3033-8892"},"institutions":[],"countries":[],"is_corresponding":true,"raw_author_name":"Siyue Zhang","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5037981788","display_name":"Yilun Zhao","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yilun Zhao","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5083573944","display_name":"Liyuan Geng","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Liyuan Geng","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5064858748","display_name":"Arman Cohan","orcid":"https://orcid.org/0000-0002-8954-2724"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Arman Cohan","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":null,"display_name":"Anh Tuan Luu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Anh Tuan Luu","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5100352016","display_name":"Zhao Chen","orcid":"https://orcid.org/0000-0003-4876-757X"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Chen Zhao","raw_affiliation_strings":[],"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":6,"corresponding_author_ids":["https://openalex.org/A5101897994"],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":true,"cited_by_count":0,"citation_normalized_percentile":{"value":0.17384095,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"4273","last_page":"4303"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T12380","display_name":"Authorship Attribution and Profiling","score":0.3395000100135803,"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/T12380","display_name":"Authorship Attribution and Profiling","score":0.3395000100135803,"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/T13910","display_name":"Computational and Text Analysis Methods","score":0.1639000028371811,"subfield":{"id":"https://openalex.org/subfields/3300","display_name":"General Social Sciences"},"field":{"id":"https://openalex.org/fields/33","display_name":"Social Sciences"},"domain":{"id":"https://openalex.org/domains/2","display_name":"Social Sciences"}},{"id":"https://openalex.org/T10028","display_name":"Topic Modeling","score":0.15139999985694885,"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/perspective","display_name":"Perspective (graphical)","score":0.6888999938964844},{"id":"https://openalex.org/keywords/embedding","display_name":"Embedding","score":0.4399999976158142},{"id":"https://openalex.org/keywords/diffusion","display_name":"Diffusion","score":0.41280001401901245},{"id":"https://openalex.org/keywords/autoregressive-model","display_name":"Autoregressive model","score":0.388700008392334},{"id":"https://openalex.org/keywords/domain","display_name":"Domain (mathematical analysis)","score":0.29899999499320984},{"id":"https://openalex.org/keywords/language-model","display_name":"Language model","score":0.2791999876499176}],"concepts":[{"id":"https://openalex.org/C12713177","wikidata":"https://www.wikidata.org/wiki/Q1900281","display_name":"Perspective (graphical)","level":2,"score":0.6888999938964844},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.5378999710083008},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.445499986410141},{"id":"https://openalex.org/C41608201","wikidata":"https://www.wikidata.org/wiki/Q980509","display_name":"Embedding","level":2,"score":0.4399999976158142},{"id":"https://openalex.org/C69357855","wikidata":"https://www.wikidata.org/wiki/Q163214","display_name":"Diffusion","level":2,"score":0.41280001401901245},{"id":"https://openalex.org/C159877910","wikidata":"https://www.wikidata.org/wiki/Q2202883","display_name":"Autoregressive model","level":2,"score":0.388700008392334},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.351500004529953},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.3497999906539917},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.3488999903202057},{"id":"https://openalex.org/C36503486","wikidata":"https://www.wikidata.org/wiki/Q11235244","display_name":"Domain (mathematical analysis)","level":2,"score":0.29899999499320984},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.2791999876499176},{"id":"https://openalex.org/C195324797","wikidata":"https://www.wikidata.org/wiki/Q33742","display_name":"Natural language","level":2,"score":0.27230000495910645},{"id":"https://openalex.org/C2779343474","wikidata":"https://www.wikidata.org/wiki/Q3109175","display_name":"Context (archaeology)","level":2,"score":0.2660999894142151},{"id":"https://openalex.org/C99498987","wikidata":"https://www.wikidata.org/wiki/Q2210247","display_name":"Noise (video)","level":3,"score":0.2630999982357025},{"id":"https://openalex.org/C80444323","wikidata":"https://www.wikidata.org/wiki/Q2878974","display_name":"Theoretical computer science","level":1,"score":0.25679999589920044},{"id":"https://openalex.org/C61797465","wikidata":"https://www.wikidata.org/wiki/Q1188986","display_name":"Term (time)","level":2,"score":0.25270000100135803},{"id":"https://openalex.org/C41895202","wikidata":"https://www.wikidata.org/wiki/Q8162","display_name":"Linguistics","level":1,"score":0.2524000108242035}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.18653/v1/2025.emnlp-main.213","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.emnlp-main.213","pdf_url":"https://aclanthology.org/2025.emnlp-main.213.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":"Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing","raw_type":"proceedings-article"}],"best_oa_location":{"id":"doi:10.18653/v1/2025.emnlp-main.213","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.emnlp-main.213","pdf_url":"https://aclanthology.org/2025.emnlp-main.213.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":"Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing","raw_type":"proceedings-article"},"sustainable_development_goals":[],"awards":[],"funders":[{"id":"https://openalex.org/F4320310501","display_name":"New York University Shanghai","ror":"https://ror.org/02vpsdb40"},{"id":"https://openalex.org/F4320320766","display_name":"Nanyang Technological University","ror":"https://ror.org/02e7b5302"}],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4416036849.pdf","grobid_xml":"https://content.openalex.org/works/W4416036849.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Large":[0],"language":[1,61,92],"model":[2,100],"(LLM)-based":[3],"embedding":[4,24,53,93,99,121],"models,":[5],"benefiting":[6],"from":[7],"large":[8],"scale":[9],"pretraining":[10],"and":[11,18,72,114,136],"post-training,":[12],"have":[13],"begun":[14],"to":[15],"surpass":[16],"BERT":[17],"T5-based":[19],"models":[20,62],"on":[21,81,103,107,111,118],"generalpurpose":[22],"text":[23,52,64,120],"tasks":[25],"such":[26],"as":[27],"document":[28],"retrieval.However,":[29],"a":[30],"fundamental":[31],"limitation":[32],"of":[33,51,89],"LLM":[34],"embeddings":[35],"lies":[36],"in":[37,75,134],"the":[38,48,85,90,97],"unidirectional":[39],"attention":[40,127],"used":[41],"during":[42],"autoregressive":[43],"pre-training,":[44],"which":[45,95],"misaligns":[46],"with":[47],"bidirectional":[49,70,126],"nature":[50],"tasks.To":[54],"this":[55],"end,":[56],"we":[57],"propose":[58],"adopting":[59],"diffusion":[60,91],"for":[63,130],"embeddings,":[65],"motivated":[66],"by":[67,101],"their":[68],"inherent":[69],"architecture":[71],"recent":[73],"success":[74],"matching":[76],"or":[77],"surpassing":[78],"LLMs":[79],"especially":[80],"reasoning":[82],"tasks.We":[83],"present":[84],"first":[86],"systematic":[87],"study":[88],"model,":[94],"outperforms":[96],"LLM-based":[98],"20%":[102],"long-document":[104],"retrieval,":[105,109,113],"8%":[106],"reasoning-intensive":[108],"2%":[110],"instruction-following":[112],"achieve":[115],"competitive":[116],"performance":[117],"traditional":[119],"benchmarks.Our":[122],"analysis":[123],"verifies":[124],"that":[125],"is":[128],"crucial":[129],"encoding":[131],"global":[132],"context":[133],"long":[135],"complex":[137],"text.":[138]},"counts_by_year":[],"updated_date":"2026-04-16T08:26:57.006410","created_date":"2025-11-08T00:00:00"}