{"id":"https://openalex.org/W7163057802","doi":"https://doi.org/10.48550/arxiv.2605.31215","title":"Fixed-Point Masked Generative Modeling","display_name":"Fixed-Point Masked Generative Modeling","publication_year":2026,"publication_date":"2026-05-29","ids":{"openalex":"https://openalex.org/W7163057802","doi":"https://doi.org/10.48550/arxiv.2605.31215"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2605.31215","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.31215","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":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"},"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.48550/arxiv.2605.31215","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5137558895","display_name":"Andrea Miele","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Miele, Andrea","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5137598904","display_name":"Yiming Qin","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Qin, Yiming","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5137568014","display_name":"Alba Carballo-Castro","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Carballo-Castro, Alba","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5137546874","display_name":"Justin Deschenaux","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Deschenaux, Justin","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5000947076","display_name":"Pascal Frossard","orcid":"https://orcid.org/0000-0002-4010-714X"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Frossard, Pascal","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":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":0.8787999749183655,"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":0.8787999749183655,"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/T10036","display_name":"Advanced Neural Network Applications","score":0.033900000154972076,"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/T11019","display_name":"Image Enhancement Techniques","score":0.01850000023841858,"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/perplexity","display_name":"Perplexity","score":0.6575999855995178},{"id":"https://openalex.org/keywords/reuse","display_name":"Reuse","score":0.557200014591217},{"id":"https://openalex.org/keywords/solver","display_name":"Solver","score":0.5462999939918518},{"id":"https://openalex.org/keywords/transformer","display_name":"Transformer","score":0.5351999998092651},{"id":"https://openalex.org/keywords/computation","display_name":"Computation","score":0.46059998869895935},{"id":"https://openalex.org/keywords/generative-model","display_name":"Generative model","score":0.43709999322891235},{"id":"https://openalex.org/keywords/generative-grammar","display_name":"Generative grammar","score":0.4296000003814697}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7616999745368958},{"id":"https://openalex.org/C100279451","wikidata":"https://www.wikidata.org/wiki/Q372193","display_name":"Perplexity","level":3,"score":0.6575999855995178},{"id":"https://openalex.org/C206588197","wikidata":"https://www.wikidata.org/wiki/Q846574","display_name":"Reuse","level":2,"score":0.557200014591217},{"id":"https://openalex.org/C2778770139","wikidata":"https://www.wikidata.org/wiki/Q1966904","display_name":"Solver","level":2,"score":0.5462999939918518},{"id":"https://openalex.org/C66322947","wikidata":"https://www.wikidata.org/wiki/Q11658","display_name":"Transformer","level":3,"score":0.5351999998092651},{"id":"https://openalex.org/C45374587","wikidata":"https://www.wikidata.org/wiki/Q12525525","display_name":"Computation","level":2,"score":0.46059998869895935},{"id":"https://openalex.org/C167966045","wikidata":"https://www.wikidata.org/wiki/Q5532625","display_name":"Generative model","level":3,"score":0.43709999322891235},{"id":"https://openalex.org/C39890363","wikidata":"https://www.wikidata.org/wiki/Q36108","display_name":"Generative grammar","level":2,"score":0.4296000003814697},{"id":"https://openalex.org/C2776436953","wikidata":"https://www.wikidata.org/wiki/Q5163215","display_name":"Consistency (knowledge bases)","level":2,"score":0.42250001430511475},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.4106999933719635},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.40709999203681946},{"id":"https://openalex.org/C57273362","wikidata":"https://www.wikidata.org/wiki/Q576722","display_name":"Decoding methods","level":2,"score":0.39800000190734863},{"id":"https://openalex.org/C2779530757","wikidata":"https://www.wikidata.org/wiki/Q1207505","display_name":"Quality (philosophy)","level":2,"score":0.3522000014781952},{"id":"https://openalex.org/C206729178","wikidata":"https://www.wikidata.org/wiki/Q2271896","display_name":"Scheduling (production processes)","level":2,"score":0.3422999978065491},{"id":"https://openalex.org/C187691185","wikidata":"https://www.wikidata.org/wiki/Q2020720","display_name":"Grid","level":2,"score":0.31540000438690186},{"id":"https://openalex.org/C23224414","wikidata":"https://www.wikidata.org/wiki/Q176769","display_name":"Hidden Markov model","level":2,"score":0.26980000734329224},{"id":"https://openalex.org/C198531522","wikidata":"https://www.wikidata.org/wiki/Q485146","display_name":"Sample (material)","level":2,"score":0.2685000002384186},{"id":"https://openalex.org/C107673813","wikidata":"https://www.wikidata.org/wiki/Q812534","display_name":"Bayesian probability","level":2,"score":0.25279998779296875}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2605.31215","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.31215","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":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2605.31215","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.31215","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":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"},"sustainable_development_goals":[{"score":0.4697979986667633,"id":"https://metadata.un.org/sdg/9","display_name":"Industry, innovation and infrastructure"}],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Masked":[0,59],"Generative":[1,60],"Models":[2,61],"(MGMs)":[3],"enable":[4,78],"parallel":[5],"decoding":[6],"and":[7,24,125,166,180,211,232],"achieve":[8],"strong":[9],"performance":[10],"across":[11],"modalities,":[12,161],"but":[13,42],"require":[14],"full-sequence":[15],"bidirectional":[16],"transformers":[17],"at":[18,103,192],"every":[19],"step,":[20],"making":[21],"training":[22,176,207,231],"costly":[23],"degrading":[25],"quality":[26,165],"under":[27],"low":[28],"sampling":[29],"budgets.":[30],"Existing":[31],"work":[32],"improves":[33,163],"efficiency":[34],"via":[35],"better":[36],"samplers":[37],"or":[38],"cheaper":[39,230],"fixed-depth":[40],"denoisers,":[41],"they":[43],"still":[44],"allocate":[45],"a":[46,70,95,193,226],"fixed":[47],"amount":[48],"of":[49,66,195],"denoiser":[50,68],"computation":[51],"to":[52,77,190,201],"each":[53],"refinement":[54],"step.":[55],"We":[56,143],"introduce":[57,94],"Fixed-Point":[58],"(FP-MGMs),":[62],"which":[63,99,112],"replace":[64],"part":[65],"the":[67,114,117,164],"with":[69,81,154],"fixed-point":[71,139],"solver":[72,115],"over":[73],"shared":[74],"attention":[75],"layers":[76],"adaptive":[79],"depth":[80],"fewer":[82],"parameters.":[83],"To":[84],"make":[85],"it":[86],"more":[87],"effective":[88],"for":[89,138,229],"masked":[90,140,235],"generation,":[91,141],"we":[92],"first":[93],"cross-step":[96],"consistency":[97],"loss,":[98],"aligns":[100],"hidden":[101],"representations":[102],"neighboring":[104],"denoising":[105],"steps":[106],"and,":[107],"second,":[108],"three-state":[109],"reuse":[110],"(3SR)":[111],"warm-starts":[113],"using":[116],"previous":[118],"solution":[119],"by":[120,174,178,182,209,213],"treating":[121],"differently":[122],"unchanged,":[123],"still-masked,":[124],"newly":[126],"revealed":[127],"tokens":[128],"respectively.":[129],"Together,":[130],"these":[131],"components":[132],"define":[133],"our":[134],"complete":[135],"training-to-inference":[136],"framework":[137,228],"\\emph{CoFRe}.":[142],"also":[144],"show":[145],"that":[146],"pre-trained":[147],"MGMs":[148],"can":[149],"be":[150],"converted":[151],"into":[152],"FP-MGMs":[153],"short":[155],"fine-tuning,":[156],"avoiding":[157],"full":[158],"retraining.":[159],"Across":[160],"CoFRe":[162,171,205,224],"cost":[167],"trade-off.":[168],"On":[169],"OpenWebText,":[170],"reduces":[172,206],"parameters":[173],"38.8\\%,":[175],"time":[177,208],"11.5\\%,":[179],"VRAM":[181,212],"16.9\\%,":[183],"while":[184,215],"improving":[185,216],"generative":[186],"perplexity":[187],"from":[188],"830.8":[189],"101.8":[191],"budget":[194],"$96$":[196],"transformer-block":[197],"forward":[198],"passes,":[199],"compared":[200],"MDLM.":[202],"In":[203],"ImageNette,":[204],"48.6\\%":[210],"50.7\\%,":[214],"FID":[217],"in":[218],"all":[219],"sample":[220],"budgets":[221],"tested.":[222],"Overall,":[223],"offers":[225],"practical":[227],"stronger":[233],"low-budget":[234],"generation.":[236]},"counts_by_year":[],"updated_date":"2026-07-01T06:00:48.157686","created_date":"2026-06-02T00:00:00"}