{"id":"https://openalex.org/W7168340407","doi":"https://doi.org/10.24963/ijcai.2026/873","title":"Accelerating Masked Diffusion Large Language Models: A Survey of Efficient Inference Techniques","display_name":"Accelerating Masked Diffusion Large Language Models: A Survey of Efficient Inference Techniques","publication_year":2026,"publication_date":"2026-09-01","ids":{"openalex":"https://openalex.org/W7168340407","doi":"https://doi.org/10.24963/ijcai.2026/873"},"language":"en","primary_location":{"id":"doi:10.24963/ijcai.2026/873","doi":"https://doi.org/10.24963/ijcai.2026/873","is_oa":true,"landing_page_url":"https://doi.org/10.24963/ijcai.2026/873","pdf_url":"https://www.ijcai.org/proceedings/2026/0873.pdf","source":{"id":"https://openalex.org/S4306419999","display_name":"International Joint Conference on Artificial Intelligence","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"conference"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the Thirty-Fifth International Joint Conference on Artificial Intelligence","raw_type":"proceedings-article"},"volume":null,"type":"conference-paper","is_traditional":true,"indexed_in":["arxiv","crossref","datacite"],"open_access":{"is_oa":true,"oa_status":"bronze","oa_url":"https://www.ijcai.org/proceedings/2026/0873.pdf","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5082233632","display_name":"Daehoon Gwak","orcid":"https://orcid.org/0009-0008-7262-1013","observed_orcids":["https://orcid.org/0009-0008-7262-1013"]},"institutions":[{"id":"https://openalex.org/I157485424","display_name":"Korea Advanced Institute of Science and Technology","ror":"https://ror.org/05apxxy63","country_code":"KR","type":"education","lineage":["https://openalex.org/I157485424"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Daehoon Gwak","raw_affiliation_strings":["Korea Advanced Institute of Science and Technology"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Korea Advanced Institute of Science and Technology","institution_ids":["https://openalex.org/I157485424"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5140771167","display_name":"Minhyung Lee","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I193775966","display_name":"Yonsei University","ror":"https://ror.org/01wjejq96","country_code":"KR","type":"education","lineage":["https://openalex.org/I193775966"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Minhyung Lee","raw_affiliation_strings":["Yonsei University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Yonsei University","institution_ids":["https://openalex.org/I193775966"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5140801348","display_name":"Junwoo Park","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I157485424","display_name":"Korea Advanced Institute of Science and Technology","ror":"https://ror.org/05apxxy63","country_code":"KR","type":"education","lineage":["https://openalex.org/I157485424"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Junwoo Park","raw_affiliation_strings":["Korea Advanced Institute of Science and Technology"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Korea Advanced Institute of Science and Technology","institution_ids":["https://openalex.org/I157485424"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5145875510","display_name":"Jaegul Choo","orcid":"https://orcid.org/0000-0003-1071-4835","observed_orcids":["https://orcid.org/0000-0003-1071-4835"]},"institutions":[{"id":"https://openalex.org/I157485424","display_name":"Korea Advanced Institute of Science and Technology","ror":"https://ror.org/05apxxy63","country_code":"KR","type":"education","lineage":["https://openalex.org/I157485424"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Jaegul Choo","raw_affiliation_strings":["Korea Advanced Institute of Science and Technology"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Korea Advanced Institute of Science and Technology","institution_ids":["https://openalex.org/I157485424"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":2,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":1.3586,"has_fulltext":true,"cited_by_count":1,"citation_normalized_percentile":{"value":0.89120531,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":91,"max":95},"biblio":{"volume":null,"issue":null,"first_page":"7864","last_page":"7872"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10181","display_name":"Natural Language Processing Techniques","score":0.5619999766349792,"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":0.5619999766349792,"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":0.22769999504089355,"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/T10036","display_name":"Advanced Neural Network Applications","score":0.06669999659061432,"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"}}],"keywords":[{"id":"https://openalex.org/keywords/diffusion-large-language-models","display_name":"diffusion large language models","score":0.9290000200271606},{"id":"https://openalex.org/keywords/inference-efficiency","display_name":"inference efficiency","score":0.7639999985694885},{"id":"https://openalex.org/keywords/parallel-generation","display_name":"parallel generation","score":0.7540000081062317},{"id":"https://openalex.org/keywords/inference-time-scaling","display_name":"inference-time scaling","score":0.6549999713897705},{"id":"https://openalex.org/keywords/large-language-model-deployment","display_name":"large language model deployment","score":0.6150000095367432}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8047999739646912},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.7957000136375427},{"id":"https://openalex.org/C130440534","wikidata":"https://www.wikidata.org/wiki/Q14946528","display_name":"Conflation","level":2,"score":0.6873000264167786},{"id":"https://openalex.org/C86251818","wikidata":"https://www.wikidata.org/wiki/Q816754","display_name":"Benchmarking","level":2,"score":0.5684999823570251},{"id":"https://openalex.org/C117896860","wikidata":"https://www.wikidata.org/wiki/Q11376","display_name":"Acceleration","level":2,"score":0.45489999651908875},{"id":"https://openalex.org/C94124525","wikidata":"https://www.wikidata.org/wiki/Q912550","display_name":"Categorization","level":2,"score":0.44130000472068787},{"id":"https://openalex.org/C82876162","wikidata":"https://www.wikidata.org/wiki/Q17096504","display_name":"Latency (audio)","level":2,"score":0.3912000060081482},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.3515999913215637},{"id":"https://openalex.org/C134261354","wikidata":"https://www.wikidata.org/wiki/Q938438","display_name":"Statistical inference","level":2,"score":0.34529998898506165},{"id":"https://openalex.org/C2779793024","wikidata":"https://www.wikidata.org/wiki/Q17299941","display_name":"Indirect Inference","level":3,"score":0.3319000005722046},{"id":"https://openalex.org/C80444323","wikidata":"https://www.wikidata.org/wiki/Q2878974","display_name":"Theoretical computer science","level":1,"score":0.3197999894618988},{"id":"https://openalex.org/C2780598303","wikidata":"https://www.wikidata.org/wiki/Q65921492","display_name":"Flexibility (engineering)","level":2,"score":0.31839999556541443},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.3167000114917755},{"id":"https://openalex.org/C46637626","wikidata":"https://www.wikidata.org/wiki/Q6693015","display_name":"Low latency (capital markets)","level":2,"score":0.31630000472068787},{"id":"https://openalex.org/C75165309","wikidata":"https://www.wikidata.org/wiki/Q2258979","display_name":"Search engine indexing","level":2,"score":0.31279999017715454},{"id":"https://openalex.org/C124681953","wikidata":"https://www.wikidata.org/wiki/Q339062","display_name":"Decomposition","level":2,"score":0.30820000171661377},{"id":"https://openalex.org/C159877910","wikidata":"https://www.wikidata.org/wiki/Q2202883","display_name":"Autoregressive model","level":2,"score":0.2628999948501587},{"id":"https://openalex.org/C3018263672","wikidata":"https://www.wikidata.org/wiki/Q1296251","display_name":"Efficient algorithm","level":2,"score":0.25589999556541443},{"id":"https://openalex.org/C63479239","wikidata":"https://www.wikidata.org/wiki/Q7353546","display_name":"Robustness (evolution)","level":3,"score":0.25189998745918274},{"id":"https://openalex.org/C2777472644","wikidata":"https://www.wikidata.org/wiki/Q16968992","display_name":"Approximate inference","level":3,"score":0.25060001015663147}],"mesh":[],"locations_count":3,"locations":[{"id":"doi:10.24963/ijcai.2026/873","doi":"https://doi.org/10.24963/ijcai.2026/873","is_oa":true,"landing_page_url":"https://doi.org/10.24963/ijcai.2026/873","pdf_url":"https://www.ijcai.org/proceedings/2026/0873.pdf","source":{"id":"https://openalex.org/S4306419999","display_name":"International Joint Conference on Artificial Intelligence","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"conference"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the Thirty-Fifth International Joint Conference on Artificial Intelligence","raw_type":"proceedings-article"},{"id":"pmh:oai:arXiv.org:2607.12829","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2607.12829","pdf_url":"https://arxiv.org/pdf/2607.12829","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"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":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"doi:10.48550/arxiv.2607.12829","doi":"https://doi.org/10.48550/arxiv.2607.12829","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2607.12829","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"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":"Preprint"}],"best_oa_location":{"id":"doi:10.24963/ijcai.2026/873","doi":"https://doi.org/10.24963/ijcai.2026/873","is_oa":true,"landing_page_url":"https://doi.org/10.24963/ijcai.2026/873","pdf_url":"https://www.ijcai.org/proceedings/2026/0873.pdf","source":{"id":"https://openalex.org/S4306419999","display_name":"International Joint Conference on Artificial Intelligence","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"conference"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the Thirty-Fifth International Joint Conference on Artificial Intelligence","raw_type":"proceedings-article"},"sustainable_development_goals":[],"study_designs":[],"awards":[{"id":"https://openalex.org/G3736587842","display_name":null,"funder_award_id":"RS-2019-II190075","funder_id":"https://openalex.org/F4320335489","funder_display_name":"Institute for Information and Communications Technology Promotion"},{"id":"https://openalex.org/G4468082379","display_name":null,"funder_award_id":"RS-2019-II190075","funder_id":"https://openalex.org/F4320328359","funder_display_name":"Ministry of Science and ICT, South Korea"},{"id":"https://openalex.org/G8705040809","display_name":null,"funder_award_id":"RS-2025-02304967","funder_id":"https://openalex.org/F4320328359","funder_display_name":"Ministry of Science and ICT, South Korea"}],"funders":[{"id":"https://openalex.org/F4320324161","display_name":"Korea Advanced Institute of Science and Technology","ror":"https://ror.org/05apxxy63"},{"id":"https://openalex.org/F4320328359","display_name":"Ministry of Science and ICT, South Korea","ror":"https://ror.org/01wpjm123"},{"id":"https://openalex.org/F4320335489","display_name":"Institute for Information and Communications Technology Promotion","ror":"https://ror.org/01g0hqq23"}],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7168340407.pdf","grobid_xml":"https://content.openalex.org/works/W7168340407.grobid-xml"},"referenced_works_count":2,"referenced_works":["https://openalex.org/W4417100850","https://openalex.org/W4417357146"],"related_works":["https://openalex.org/W7128788044","https://openalex.org/W7090315770","https://openalex.org/W7161116509","https://openalex.org/W7125531142","https://openalex.org/W7135044728","https://openalex.org/W7133238215","https://openalex.org/W7166199774","https://openalex.org/W4416872067","https://openalex.org/W4415037641","https://openalex.org/W7128384668"],"abstract_inverted_index":{"Diffusion":[0],"large":[1],"language":[2],"models":[3],"(dLLMs)":[4],"offer":[5],"a":[6,43,90],"theoretical":[7],"advantage":[8],"in":[9,82,108],"parallel":[10,17,149],"generation":[11,18],"over":[12],"standard":[13],"autoregressive":[14],"models.":[15],"However,":[16,60],"alone":[19],"does":[20],"not":[21],"guarantee":[22],"practical":[23,46],"speedups.":[24],"Realizing":[25],"this":[26,86,113],"efficiency":[27,41],"requires":[28],"specialized":[29],"inference":[30,40,106],"mechanisms,":[31],"such":[32],"as":[33,39,65],"diffusion-aware":[34],"caching":[35],"and":[36,58,75,101,126,129,139],"reuse.":[37],"Consequently,":[38],"becomes":[42],"prerequisite":[44],"for":[45,95,136,143],"deployment,":[47],"recent":[48],"research":[49],"has":[50],"actively":[51],"explored":[52],"acceleration":[53,117],"techniques":[54,118],"across":[55],"algorithms,":[56],"architectures,":[57],"systems.":[59],"rigorous":[61],"comparisons":[62],"remain":[63],"difficult,":[64],"end-to-end":[66],"latency":[67,92],"stems":[68],"from":[69],"intricate":[70],"trade-offs":[71],"between":[72],"algorithmic,":[73],"architectural,":[74],"system-level":[76],"factors":[77,100],"that":[78],"are":[79],"often":[80],"conflated":[81],"existing":[83],"benchmarks.":[84],"In":[85],"survey,":[87],"we":[88,115,133],"introduce":[89],"unified":[91],"decomposition":[93],"framework":[94],"dLLMs":[96],"to":[97],"disentangle":[98],"these":[99],"analyze":[102],"their":[103],"impact":[104],"on":[105],"speed":[107],"real":[109],"deployments.":[110],"Guided":[111],"by":[112],"framework,":[114],"categorize":[116],"along":[119],"three":[120],"axes":[121],"covering":[122],"algorithmic":[123],"innovations,":[124],"architectural":[125],"system":[127],"optimizations,":[128],"inference-time":[130],"scaling.":[131],"Finally,":[132],"provide":[134],"guidelines":[135],"reproducible":[137],"benchmarking":[138],"highlight":[140],"open":[141],"challenges":[142],"realizing":[144],"the":[145],"full":[146],"potential":[147],"of":[148],"generation.":[150]},"counts_by_year":[{"year":2026,"cited_by_count":1}],"updated_date":"2026-11-07T11:25:07.637180","created_date":"2026-07-16T00:00:00"}