{"id":"https://openalex.org/W7154266495","doi":"https://doi.org/10.48550/arxiv.2604.11035","title":"Introspective Diffusion Language Models","display_name":"Introspective Diffusion Language Models","publication_year":2026,"publication_date":"2026-04-13","ids":{"openalex":"https://openalex.org/W7154266495","doi":"https://doi.org/10.48550/arxiv.2604.11035"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2604.11035","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.11035","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"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":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"article"},"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.48550/arxiv.2604.11035","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5133576886","display_name":"Yifan Yu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yu, Yifan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5133371582","display_name":"Yuqing Jian","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Jian, Yuqing","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5133585570","display_name":"Junxiong Wang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wang, Junxiong","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5133555764","display_name":"Zhongzhu Zhou","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhou, Zhongzhu","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5051047559","display_name":"Donglin Zhuang","orcid":"https://orcid.org/0000-0003-3355-407X"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhuang, Donglin","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5133554246","display_name":"Xinyu Fang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Fang, Xinyu","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5125406207","display_name":"Sri Yanamandra","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yanamandra, Sri","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5133566191","display_name":"Xiaoxia Wu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wu, Xiaoxia","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5133582090","display_name":"Qingyang Wu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wu, Qingyang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5133576030","display_name":"Shuaiwen Leon Song","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Song, Shuaiwen Leon","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5133608979","display_name":"Tri Dao","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Dao, Tri","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5133589273","display_name":"Ben Athiwaratkun","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Athiwaratkun, Ben","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5133574962","display_name":"James Zou","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zou, James","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5133611714","display_name":"Fan Lai","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Lai, Fan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5133604825","display_name":"Chenfeng Xu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Xu, Chenfeng","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":15,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":null,"last_page":null},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10775","display_name":"Generative Adversarial Networks and Image Synthesis","score":0.5504999756813049,"subfield":{"id":"https://openalex.org/subfields/1707","display_name":"Computer Vision and Pattern Recognition"},"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/T10775","display_name":"Generative Adversarial Networks and Image Synthesis","score":0.5504999756813049,"subfield":{"id":"https://openalex.org/subfields/1707","display_name":"Computer Vision and Pattern Recognition"},"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/T11307","display_name":"Domain Adaptation and Few-Shot Learning","score":0.07370000332593918,"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":0.06419999897480011,"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/introspection","display_name":"Introspection","score":0.8597999811172485},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.5199999809265137},{"id":"https://openalex.org/keywords/quality","display_name":"Quality (philosophy)","score":0.5005000233650208},{"id":"https://openalex.org/keywords/language-model","display_name":"Language model","score":0.5002999901771545},{"id":"https://openalex.org/keywords/decoding-methods","display_name":"Decoding methods","score":0.46309998631477356},{"id":"https://openalex.org/keywords/consistency","display_name":"Consistency (knowledge bases)","score":0.42559999227523804},{"id":"https://openalex.org/keywords/autoregressive-model","display_name":"Autoregressive model","score":0.36500000953674316},{"id":"https://openalex.org/keywords/diffusion","display_name":"Diffusion","score":0.3578000068664551}],"concepts":[{"id":"https://openalex.org/C129671850","wikidata":"https://www.wikidata.org/wiki/Q210501","display_name":"Introspection","level":2,"score":0.8597999811172485},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7714999914169312},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.5199999809265137},{"id":"https://openalex.org/C2779530757","wikidata":"https://www.wikidata.org/wiki/Q1207505","display_name":"Quality (philosophy)","level":2,"score":0.5005000233650208},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.5002999901771545},{"id":"https://openalex.org/C57273362","wikidata":"https://www.wikidata.org/wiki/Q576722","display_name":"Decoding methods","level":2,"score":0.46309998631477356},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.4521999955177307},{"id":"https://openalex.org/C2776436953","wikidata":"https://www.wikidata.org/wiki/Q5163215","display_name":"Consistency (knowledge bases)","level":2,"score":0.42559999227523804},{"id":"https://openalex.org/C80444323","wikidata":"https://www.wikidata.org/wiki/Q2878974","display_name":"Theoretical computer science","level":1,"score":0.40299999713897705},{"id":"https://openalex.org/C159877910","wikidata":"https://www.wikidata.org/wiki/Q2202883","display_name":"Autoregressive model","level":2,"score":0.36500000953674316},{"id":"https://openalex.org/C69357855","wikidata":"https://www.wikidata.org/wiki/Q163214","display_name":"Diffusion","level":2,"score":0.3578000068664551},{"id":"https://openalex.org/C120524526","wikidata":"https://www.wikidata.org/wiki/Q1709148","display_name":"Reboot","level":2,"score":0.34049999713897705},{"id":"https://openalex.org/C125411270","wikidata":"https://www.wikidata.org/wiki/Q18653","display_name":"Encoding (memory)","level":2,"score":0.33090001344680786},{"id":"https://openalex.org/C2780009758","wikidata":"https://www.wikidata.org/wiki/Q6804172","display_name":"Measure (data warehouse)","level":2,"score":0.310699999332428},{"id":"https://openalex.org/C2781238097","wikidata":"https://www.wikidata.org/wiki/Q175026","display_name":"Object (grammar)","level":2,"score":0.3041999936103821},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.29989999532699585},{"id":"https://openalex.org/C2780801425","wikidata":"https://www.wikidata.org/wiki/Q5164392","display_name":"Construct (python library)","level":2,"score":0.2973000109195709},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.2897999882698059},{"id":"https://openalex.org/C2777402240","wikidata":"https://www.wikidata.org/wiki/Q6783436","display_name":"Masking (illustration)","level":2,"score":0.2831000089645386},{"id":"https://openalex.org/C11671645","wikidata":"https://www.wikidata.org/wiki/Q5054567","display_name":"Causal model","level":2,"score":0.27480000257492065},{"id":"https://openalex.org/C100279451","wikidata":"https://www.wikidata.org/wiki/Q372193","display_name":"Perplexity","level":3,"score":0.2709999978542328},{"id":"https://openalex.org/C2780069185","wikidata":"https://www.wikidata.org/wiki/Q7977945","display_name":"Equivalence (formal languages)","level":2,"score":0.27059999108314514},{"id":"https://openalex.org/C66024118","wikidata":"https://www.wikidata.org/wiki/Q1122506","display_name":"Computational model","level":2,"score":0.2596000134944916},{"id":"https://openalex.org/C81917197","wikidata":"https://www.wikidata.org/wiki/Q628760","display_name":"Selection (genetic algorithm)","level":2,"score":0.2529999911785126}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2604.11035","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.11035","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"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":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"article"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2604.11035","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.11035","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"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":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"article"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Diffusion":[0,78],"language":[1],"models":[2,12,26],"promise":[3],"parallel":[4,87],"generation,":[5],"yet":[6],"still":[7],"lag":[8],"behind":[9],"autoregressive":[10],"(AR)":[11],"in":[13,119,168],"quality.":[14],"We":[15,37],"stem":[16],"this":[17,73],"gap":[18],"to":[19,110,155],"a":[20,46,59,82,99,125,141],"failure":[21],"of":[22,94,147,159,208],"introspective":[23,40,69,92,101],"consistency:":[24],"AR":[25,56,95,162],"agree":[27],"with":[28,140],"their":[29],"own":[30],"generations,":[31],"while":[32,89,115,164],"DLMs":[33,167],"often":[34],"do":[35],"not.":[36],"define":[38],"the":[39,91,108,120,145,152,157,205],"acceptance":[41],"rate,":[42],"which":[43,106],"measures":[44],"whether":[45],"model":[47,109,170],"accepts":[48],"its":[49,160],"previously":[50,112],"generated":[51,113],"tokens.":[52],"This":[53],"reveals":[54],"why":[55],"training":[57],"has":[58],"structural":[60],"advantage:":[61],"causal":[62],"masking":[63],"and":[64,136,172,184,195],"logit":[65],"shifting":[66],"implicitly":[67],"enforce":[68],"consistency.":[70],"Motivated":[71],"by":[72,191],"observation,":[74],"we":[75,128],"introduce":[76],"Introspective":[77],"Language":[79],"Model":[80],"(I-DLM),":[81],"paradigm":[83],"that":[84],"retains":[85],"diffusion-style":[86],"decoding":[88,103],"inheriting":[90],"consistency":[93],"training.":[96],"I-DLM":[97,130,150,201],"uses":[98],"novel":[100],"strided":[102],"(ISD)":[104],"algorithm,":[105],"enables":[107],"verify":[111],"tokens":[114],"advancing":[116],"new":[117],"ones":[118],"same":[121],"forward":[122],"pass.":[123],"From":[124],"systems":[126],"standpoint,":[127],"build":[129],"inference":[131],"engine":[132],"on":[133,182,186],"AR-inherited":[134],"optimizations":[135],"further":[137],"customize":[138],"it":[139],"stationary-batch":[142],"scheduler.":[143],"To":[144],"best":[146],"our":[148],"knowledge,":[149],"is":[151,202],"first":[153],"DLM":[154],"match":[156],"quality":[158,171],"same-scale":[161],"counterpart":[163],"outperforming":[165],"prior":[166,217],"both":[169],"practical":[173],"serving":[174],"efficiency":[175],"across":[176],"15":[177,196],"benchmarks.":[178],"It":[179],"reaches":[180],"69.6":[181],"AIME-24":[183],"45.7":[185],"LiveCodeBench-v6,":[187],"exceeding":[188],"LLaDA-2.1-mini":[189],"(16B)":[190],"more":[192],"than":[193,216],"26":[194],"points,":[197],"respectively.":[198],"Beyond":[199],"quality,":[200],"designed":[203],"for":[204],"growing":[206],"demand":[207],"large-concurrency":[209],"serving,":[210],"delivering":[211],"about":[212],"3x":[213],"higher":[214],"throughput":[215],"state-of-the-art":[218],"DLMs.":[219]},"counts_by_year":[],"updated_date":"2026-06-11T09:08:48.828518","created_date":"2026-04-15T00:00:00"}