{"id":"https://openalex.org/W7224125183","doi":"https://doi.org/10.48550/arxiv.2610.37642","title":"Metacognitive Decoding: Efficient Inference for Diffusion Language Agents","display_name":"Metacognitive Decoding: Efficient Inference for Diffusion Language Agents","publication_year":2026,"publication_date":"2026-10-27","ids":{"openalex":"https://openalex.org/W7224125183","doi":"https://doi.org/10.48550/arxiv.2610.37642"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2610.37642","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2610.37642","pdf_url":"https://arxiv.org/pdf/2610.37642","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"},"volume":null,"type":"preprint","indexed_in":["arxiv","datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/2610.37642","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5160040331","display_name":"Mingyu Zhao","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhao, Mingyu","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5160063092","display_name":"Shuhang Lin","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Lin, Shuhang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5160007230","display_name":"Ligong Han","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Han, Ligong","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5159173902","display_name":"Xiaojing Yu","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yu, Xiaojing","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5160053833","display_name":"Zhenting Wang","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wang, Zhenting","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5125368833","display_name":"Yifei Pang","orcid":"https://orcid.org/0009-0004-4515-4422","observed_orcids":["https://orcid.org/0009-0004-4515-4422"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Pang, Yifei","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5159989902","display_name":"Yang Zhou","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhou, Yang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5064437234","display_name":"Dimitris Metaxas","orcid":"https://orcid.org/0000-0001-7142-7640","observed_orcids":["https://orcid.org/0000-0001-7142-7640"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Metaxas, Dimitris N.","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":null,"has_fulltext":true,"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/T10181","display_name":"Natural Language Processing Techniques","score":0.538100004196167,"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":0.538100004196167,"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":0.2409999966621399,"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/T11714","display_name":"Multimodal Machine Learning Applications","score":0.0414000004529953,"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-language-models","display_name":"diffusion language models","score":0.8690000176429749},{"id":"https://openalex.org/keywords/metacognitive-control","display_name":"metacognitive regulation","score":0.8190000057220459},{"id":"https://openalex.org/keywords/token-generation","display_name":"token generation","score":0.41499999165534973},{"id":"https://openalex.org/keywords/inference-efficiency","display_name":"inference efficiency","score":0.7789999842643738},{"id":"https://openalex.org/keywords/generation-cost-reduction","display_name":"generation cost reduction","score":0.5960000157356262},{"id":"https://openalex.org/keywords/language-agents","display_name":"language agents","score":0.5920000076293945},{"id":"https://openalex.org/keywords/flops-reduction","display_name":"FLOPs reduction","score":0.5360000133514404}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8237000107765198},{"id":"https://openalex.org/C48145219","wikidata":"https://www.wikidata.org/wiki/Q1335365","display_name":"Security token","level":2,"score":0.6776000261306763},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.6460000276565552},{"id":"https://openalex.org/C45374587","wikidata":"https://www.wikidata.org/wiki/Q12525525","display_name":"Computation","level":2,"score":0.6378999948501587},{"id":"https://openalex.org/C55439883","wikidata":"https://www.wikidata.org/wiki/Q360812","display_name":"Correctness","level":2,"score":0.5720000267028809},{"id":"https://openalex.org/C98045186","wikidata":"https://www.wikidata.org/wiki/Q205663","display_name":"Process (computing)","level":2,"score":0.5559999942779541},{"id":"https://openalex.org/C2780451532","wikidata":"https://www.wikidata.org/wiki/Q759676","display_name":"Task (project management)","level":2,"score":0.5198000073432922},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.4832000136375427},{"id":"https://openalex.org/C165064840","wikidata":"https://www.wikidata.org/wiki/Q1321061","display_name":"Matching (statistics)","level":2,"score":0.4101000130176544},{"id":"https://openalex.org/C69357855","wikidata":"https://www.wikidata.org/wiki/Q163214","display_name":"Diffusion","level":2,"score":0.4016999900341034},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.373199999332428},{"id":"https://openalex.org/C157764524","wikidata":"https://www.wikidata.org/wiki/Q1383412","display_name":"Throughput","level":3,"score":0.3528999984264374},{"id":"https://openalex.org/C2777735758","wikidata":"https://www.wikidata.org/wiki/Q817765","display_name":"Path (computing)","level":2,"score":0.31690001487731934},{"id":"https://openalex.org/C203479927","wikidata":"https://www.wikidata.org/wiki/Q5165939","display_name":"Controller (irrigation)","level":2,"score":0.3046000003814697},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.30410000681877136},{"id":"https://openalex.org/C118147538","wikidata":"https://www.wikidata.org/wiki/Q1126970","display_name":"Metacognition","level":3,"score":0.29019999504089355},{"id":"https://openalex.org/C106189395","wikidata":"https://www.wikidata.org/wiki/Q176789","display_name":"Markov decision process","level":3,"score":0.28220000863075256},{"id":"https://openalex.org/C175154964","wikidata":"https://www.wikidata.org/wiki/Q380077","display_name":"Task analysis","level":3,"score":0.2799000144004822},{"id":"https://openalex.org/C12713177","wikidata":"https://www.wikidata.org/wiki/Q1900281","display_name":"Perspective (graphical)","level":2,"score":0.27639999985694885},{"id":"https://openalex.org/C13662910","wikidata":"https://www.wikidata.org/wiki/Q193139","display_name":"Trajectory","level":2,"score":0.25380000472068787},{"id":"https://openalex.org/C80444323","wikidata":"https://www.wikidata.org/wiki/Q2878974","display_name":"Theoretical computer science","level":1,"score":0.2522999942302704}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2610.37642","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2610.37642","pdf_url":"https://arxiv.org/pdf/2610.37642","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.2610.37642","doi":"https://doi.org/10.48550/arxiv.2610.37642","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2610.37642","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":"pmh:oai:arXiv.org:2610.37642","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2610.37642","pdf_url":"https://arxiv.org/pdf/2610.37642","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"},"sustainable_development_goals":[],"study_designs":[],"awards":[],"funders":[],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7224125183.pdf","grobid_xml":"https://content.openalex.org/works/W7224125183.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":["https://openalex.org/W7204194770","https://openalex.org/W4320724720","https://openalex.org/W7224137989","https://openalex.org/W7128467652","https://openalex.org/W7134247832","https://openalex.org/W7128681798","https://openalex.org/W4417448099","https://openalex.org/W7163024699","https://openalex.org/W4415037641","https://openalex.org/W7125531142"],"abstract_inverted_index":{"Diffusion":[0],"language":[1],"models":[2],"offer":[3],"a":[4,32,73],"promising":[5],"path":[6],"to":[7,40,57,85,98],"efficient":[8,15,123],"agents":[9],"through":[10],"parallel":[11],"token":[12],"generation.":[13],"However,":[14],"interaction":[16,50,100],"requires":[17],"adapting":[18],"both":[19],"where":[20],"computation":[21,68],"is":[22,28,80],"spent":[23],"and":[24,47,60,88],"how":[25,61],"much":[26],"text":[27],"generated.":[29],"We":[30],"introduce":[31],"metacognitive":[33],"controller":[34],"that":[35,109],"guides":[36],"diffusion":[37,78],"generation":[38,63,89,96],"according":[39],"the":[41,44,48,52,77,94,99,101,121,134],"needs":[42],"of":[43,136],"current":[45],"response":[46],"agent's":[49,95],"with":[51,120],"environment.":[53],"It":[54],"learns":[55],"when":[56],"finalize":[58],"tokens":[59],"far":[62],"should":[64],"extend,":[65],"reducing":[66],"overall":[67],"within":[69],"responses.":[70],"Our":[71],"controller,":[72],"lightweight":[74],"plug-in":[75],"for":[76],"backbone,":[79],"trained":[81],"on":[82,117],"agent":[83,102],"trajectories":[84],"balance":[86],"correctness":[87],"cost.":[90],"This":[91],"approach":[92,111],"connects":[93],"process":[97],"serves.":[103],"Experiments":[104],"across":[105],"five":[106],"benchmarks":[107],"demonstrate":[108],"our":[110,137],"reduces":[112],"inference":[113],"FLOPs":[114],"by":[115],"20.6%":[116],"average":[118],"compared":[119],"most":[122],"baseline,":[124],"while":[125],"maintaining":[126],"competitive":[127],"task":[128],"accuracy.":[129],"Ablation":[130],"studies":[131],"further":[132],"validate":[133],"effectiveness":[135],"approach.":[138]},"counts_by_year":[],"updated_date":"2026-10-31T12:09:11.096373","created_date":"2026-10-29T00:00:00"}