{"id":"https://openalex.org/W4417129363","doi":"https://doi.org/10.18653/v1/2026.findings-acl.170","title":"Mask Tokens as Prophet: Fine-Grained Cache Eviction for Efficient dLLM Inference","display_name":"Mask Tokens as Prophet: Fine-Grained Cache Eviction for Efficient dLLM Inference","publication_year":2026,"publication_date":"2026-01-01","ids":{"openalex":"https://openalex.org/W4417129363","doi":"https://doi.org/10.18653/v1/2026.findings-acl.170"},"language":"en","primary_location":{"id":"doi:10.18653/v1/2026.findings-acl.170","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.findings-acl.170","pdf_url":"https://aclanthology.org/2026.findings-acl.170.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":["arxiv","crossref","datacite"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://aclanthology.org/2026.findings-acl.170.pdf","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5143173573","display_name":"Jianuo Huang","orcid":null},"institutions":[{"id":"https://openalex.org/I183067930","display_name":"Shanghai Jiao Tong University","ror":"https://ror.org/0220qvk04","country_code":"CN","type":"education","lineage":["https://openalex.org/I183067930"]},{"id":"https://openalex.org/I4210164862","display_name":"Artificial Intelligence in Medicine (Canada)","ror":"https://ror.org/05p590m36","country_code":"CA","type":"company","lineage":["https://openalex.org/I4210164862"]},{"id":"https://openalex.org/I47720641","display_name":"Huazhong University of Science and Technology","ror":"https://ror.org/00p991c53","country_code":"CN","type":"education","lineage":["https://openalex.org/I47720641"]}],"countries":["CA","CN"],"is_corresponding":false,"raw_author_name":"Jianuo Huang","raw_affiliation_strings":["Huazhong University of Science and Technology","School of Artificial Intelligence , Shanghai Jiao Tong University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Huazhong University of Science and Technology","institution_ids":["https://openalex.org/I47720641"]},{"raw_affiliation_string":"School of Artificial Intelligence , Shanghai Jiao Tong University","institution_ids":["https://openalex.org/I183067930","https://openalex.org/I4210164862"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5002643603","display_name":"Yaojie Zhang","orcid":"https://orcid.org/0000-0002-4220-1623"},"institutions":[{"id":"https://openalex.org/I150229711","display_name":"University of Electronic Science and Technology of China","ror":"https://ror.org/04qr3zq92","country_code":"CN","type":"education","lineage":["https://openalex.org/I150229711"]},{"id":"https://openalex.org/I183067930","display_name":"Shanghai Jiao Tong University","ror":"https://ror.org/0220qvk04","country_code":"CN","type":"education","lineage":["https://openalex.org/I183067930"]},{"id":"https://openalex.org/I4210164862","display_name":"Artificial Intelligence in Medicine (Canada)","ror":"https://ror.org/05p590m36","country_code":"CA","type":"company","lineage":["https://openalex.org/I4210164862"]}],"countries":["CA","CN"],"is_corresponding":false,"raw_author_name":"Yaojie Zhang","raw_affiliation_strings":["School of Artificial Intelligence , Shanghai Jiao Tong University","University of Electronic Science and Technology of China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"School of Artificial Intelligence , Shanghai Jiao Tong University","institution_ids":["https://openalex.org/I183067930","https://openalex.org/I4210164862"]},{"raw_affiliation_string":"University of Electronic Science and Technology of China","institution_ids":["https://openalex.org/I150229711"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5081068266","display_name":"Yicun Yang","orcid":null},"institutions":[{"id":"https://openalex.org/I183067930","display_name":"Shanghai Jiao Tong University","ror":"https://ror.org/0220qvk04","country_code":"CN","type":"education","lineage":["https://openalex.org/I183067930"]},{"id":"https://openalex.org/I4210164862","display_name":"Artificial Intelligence in Medicine (Canada)","ror":"https://ror.org/05p590m36","country_code":"CA","type":"company","lineage":["https://openalex.org/I4210164862"]}],"countries":["CA","CN"],"is_corresponding":false,"raw_author_name":"Yicun Yang","raw_affiliation_strings":["School of Artificial Intelligence , Shanghai Jiao Tong University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"School of Artificial Intelligence , Shanghai Jiao Tong University","institution_ids":["https://openalex.org/I183067930","https://openalex.org/I4210164862"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101298780","display_name":"Benhao Huang","orcid":null},"institutions":[{"id":"https://openalex.org/I74973139","display_name":"Carnegie Mellon University","ror":"https://ror.org/05x2bcf33","country_code":"US","type":"education","lineage":["https://openalex.org/I74973139"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Benhao Huang","raw_affiliation_strings":["Carnegie Mellon University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Carnegie Mellon University","institution_ids":["https://openalex.org/I74973139"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5143292636","display_name":"Linfeng Zhang","orcid":null},"institutions":[{"id":"https://openalex.org/I183067930","display_name":"Shanghai Jiao Tong University","ror":"https://ror.org/0220qvk04","country_code":"CN","type":"education","lineage":["https://openalex.org/I183067930"]},{"id":"https://openalex.org/I4210164862","display_name":"Artificial Intelligence in Medicine (Canada)","ror":"https://ror.org/05p590m36","country_code":"CA","type":"company","lineage":["https://openalex.org/I4210164862"]}],"countries":["CA","CN"],"is_corresponding":false,"raw_author_name":"Linfeng Zhang","raw_affiliation_strings":["School of Artificial Intelligence , Shanghai Jiao Tong University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"School of Artificial Intelligence , Shanghai Jiao Tong University","institution_ids":["https://openalex.org/I183067930","https://openalex.org/I4210164862"]}]}],"institutions":[],"countries_distinct_count":3,"institutions_distinct_count":5,"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":0.03442875,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"3456","last_page":"3479"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10028","display_name":"Topic Modeling","score":0.1111999973654747,"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/T10028","display_name":"Topic Modeling","score":0.1111999973654747,"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":0.10010000318288803,"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/T10181","display_name":"Natural Language Processing Techniques","score":0.09749999642372131,"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/cache","display_name":"Cache","score":0.8398000001907349},{"id":"https://openalex.org/keywords/cache-algorithms","display_name":"Cache algorithms","score":0.6082000136375427},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.512499988079071},{"id":"https://openalex.org/keywords/cache-invalidation","display_name":"Cache invalidation","score":0.5098999738693237},{"id":"https://openalex.org/keywords/cache-coloring","display_name":"Cache coloring","score":0.503000020980835},{"id":"https://openalex.org/keywords/cache-pollution","display_name":"Cache pollution","score":0.49959999322891235},{"id":"https://openalex.org/keywords/code","display_name":"Code (set theory)","score":0.4812000095844269},{"id":"https://openalex.org/keywords/key","display_name":"Key (lock)","score":0.4375}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8603000044822693},{"id":"https://openalex.org/C115537543","wikidata":"https://www.wikidata.org/wiki/Q165596","display_name":"Cache","level":2,"score":0.8398000001907349},{"id":"https://openalex.org/C38556500","wikidata":"https://www.wikidata.org/wiki/Q13404475","display_name":"Cache algorithms","level":4,"score":0.6082000136375427},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.512499988079071},{"id":"https://openalex.org/C25536678","wikidata":"https://www.wikidata.org/wiki/Q5015977","display_name":"Cache invalidation","level":5,"score":0.5098999738693237},{"id":"https://openalex.org/C201148951","wikidata":"https://www.wikidata.org/wiki/Q5015976","display_name":"Cache coloring","level":4,"score":0.503000020980835},{"id":"https://openalex.org/C113166858","wikidata":"https://www.wikidata.org/wiki/Q5015981","display_name":"Cache pollution","level":5,"score":0.49959999322891235},{"id":"https://openalex.org/C2776760102","wikidata":"https://www.wikidata.org/wiki/Q5139990","display_name":"Code (set theory)","level":3,"score":0.4812000095844269},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.4731999933719635},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.4375},{"id":"https://openalex.org/C57273362","wikidata":"https://www.wikidata.org/wiki/Q576722","display_name":"Decoding methods","level":2,"score":0.42730000615119934},{"id":"https://openalex.org/C120314980","wikidata":"https://www.wikidata.org/wiki/Q180634","display_name":"Distributed computing","level":1,"score":0.4154999852180481},{"id":"https://openalex.org/C167713795","wikidata":"https://www.wikidata.org/wiki/Q352090","display_name":"Smart Cache","level":5,"score":0.3962000012397766},{"id":"https://openalex.org/C31258907","wikidata":"https://www.wikidata.org/wiki/Q1301371","display_name":"Computer network","level":1,"score":0.3716999888420105},{"id":"https://openalex.org/C189783530","wikidata":"https://www.wikidata.org/wiki/Q352090","display_name":"CPU cache","level":3,"score":0.3652999997138977},{"id":"https://openalex.org/C48145219","wikidata":"https://www.wikidata.org/wiki/Q1335365","display_name":"Security token","level":2,"score":0.3619999885559082},{"id":"https://openalex.org/C77618280","wikidata":"https://www.wikidata.org/wiki/Q1155772","display_name":"Scheme (mathematics)","level":2,"score":0.33709999918937683},{"id":"https://openalex.org/C45374587","wikidata":"https://www.wikidata.org/wiki/Q12525525","display_name":"Computation","level":2,"score":0.3075999915599823},{"id":"https://openalex.org/C79403827","wikidata":"https://www.wikidata.org/wiki/Q3988","display_name":"Real-time computing","level":1,"score":0.29789999127388},{"id":"https://openalex.org/C186967261","wikidata":"https://www.wikidata.org/wiki/Q5082128","display_name":"Mobile device","level":2,"score":0.2784999907016754},{"id":"https://openalex.org/C125411270","wikidata":"https://www.wikidata.org/wiki/Q18653","display_name":"Encoding (memory)","level":2,"score":0.27619999647140503},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.27390000224113464},{"id":"https://openalex.org/C43126263","wikidata":"https://www.wikidata.org/wiki/Q128751","display_name":"Source code","level":2,"score":0.2680000066757202},{"id":"https://openalex.org/C2778227907","wikidata":"https://www.wikidata.org/wiki/Q1893186","display_name":"Eviction","level":2,"score":0.26510000228881836},{"id":"https://openalex.org/C113775141","wikidata":"https://www.wikidata.org/wiki/Q428691","display_name":"Computer engineering","level":1,"score":0.25999999046325684},{"id":"https://openalex.org/C93996380","wikidata":"https://www.wikidata.org/wiki/Q44127","display_name":"Server","level":2,"score":0.25850000977516174},{"id":"https://openalex.org/C59687516","wikidata":"https://www.wikidata.org/wiki/Q5015938","display_name":"Cache-oblivious algorithm","level":5,"score":0.25459998846054077}],"mesh":[],"locations_count":3,"locations":[{"id":"doi:10.18653/v1/2026.findings-acl.170","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.findings-acl.170","pdf_url":"https://aclanthology.org/2026.findings-acl.170.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"},{"id":"pmh:oai:arXiv.org:2510.09309","is_oa":true,"landing_page_url":"http://arxiv.org/abs/2510.09309","pdf_url":"https://arxiv.org/pdf/2510.09309","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,"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.2510.09309","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2510.09309","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,"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":"Preprint"}],"best_oa_location":{"id":"doi:10.18653/v1/2026.findings-acl.170","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.findings-acl.170","pdf_url":"https://aclanthology.org/2026.findings-acl.170.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/F4320316083","display_name":"Tencent","ror":"https://ror.org/00hhjss72"}],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4417129363.pdf","grobid_xml":"https://content.openalex.org/works/W4417129363.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Diffusion":[0],"large":[1,38],"language":[2],"models":[3,12],"(dLLMs)":[4],"present":[5],"a":[6,84,108],"promising":[7],"alternative":[8],"to":[9,44,61,90,117,156,176],"dominant":[10],"autoregressive":[11],"(ARMs)":[13],"by":[14,137],"the":[15,21,29,64,94,153,168],"ability":[16,43],"of":[17,23,66,96,163,167],"parallel":[18],"decoding":[19],"at":[20,180],"expense":[22],"substantial":[24],"computation":[25],"and":[26,59,69,119,143,173],"memory":[27,39],"costs.Specifically,":[28],"cache":[30,51,86,131,155],"mechanism":[31,112],"for":[32,57,63,125],"bidirectional":[33],"attention":[34,115],"in":[35,72,75,99,140],"dLLMs":[36],"demands":[37],"footprint,":[40],"restricting":[41],"their":[42],"handle":[45],"long":[46],"contexts":[47],"under":[48],"resource-limited":[49],"settings.Existing":[50],"eviction":[52,87],"strategies":[53],"are":[54],"primarily":[55],"designed":[56],"ARMs":[58],"fail":[60],"account":[62],"role":[65],"mask":[67,97],"tokens":[68,98,124],"specific":[70],"characteristics":[71],"dLLMs,":[73,91],"resulting":[74],"suboptimal":[76],"performance.To":[77],"address":[78],"these":[79],"challenges,":[80],"we":[81],"introduce":[82],"MaskKV,":[83,151],"training-free":[85],"framework":[88],"tailored":[89],"focusing":[92],"on":[93,103,146,171,188],"effect":[95],"dLLMs.MaskKV":[100],"is":[101],"built":[102],"two":[104],"key":[105],"innovations:":[106],"(1)":[107],"mask-query":[109],"guided":[110],"scoring":[111],"that":[113,134],"leverages":[114],"weights":[116],"identify":[118],"evict":[120],"less":[121],"critical":[122],"prompt":[123,182],"each":[126],"head;":[127],"(2)":[128],"an":[129],"adaptive":[130],"budgeting":[132],"strategy":[133],"improves":[135],"efficiency":[136],"reducing":[138],"allocation":[139],"intermediate":[141],"layers":[142,200],"concentrating":[144],"resources":[145],"prompt-preferring":[147],"heads.On":[148],"LLaDA":[149,209],"with":[150],"compressing":[152],"KV":[154],"only":[157],"256":[158],"pairs":[159],"(less":[160],"than":[161],"5%":[162],"tokens)":[164],"retains":[165],"94%":[166],"full-cache":[169],"performance":[170],"LongBench":[172],"achieves":[174],"up":[175],"31":[177],"\u00d7":[178],"acceleration":[179],"32k":[181],"length.Our":[183],"code":[184],"will":[185],"be":[186],"released":[187],"Github.Key":[189],"Query":[190],"(c)":[191],"Attention":[192],"map":[193],"across":[194,199],"heads":[195],"(b)":[196],"Cosine":[197],"similarity":[198],"Sample":[201,203],"Index":[202,204,206,208],"Layer":[205,207],"Dream":[210],"0.0":[211,219],"0.2":[212,220],"0.4":[213,221],"0.6":[214],"0.8":[215,217],"1.0":[216,218],"0.":[222]},"counts_by_year":[],"updated_date":"2026-08-26T07:47:46.906454","created_date":"2025-10-14T00:00:00"}