{"id":"https://openalex.org/W7164839289","doi":"https://doi.org/10.48550/arxiv.2606.14700","title":"RepFusion: Leveraging Multimodal Priors for Denoising in Representation Space","display_name":"RepFusion: Leveraging Multimodal Priors for Denoising in Representation Space","publication_year":2026,"publication_date":"2026-06-12","ids":{"openalex":"https://openalex.org/W7164839289","doi":"https://doi.org/10.48550/arxiv.2606.14700"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2606.14700","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.14700","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":"Preprint"},"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.48550/arxiv.2606.14700","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5003231318","display_name":"Xichen Pan","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Pan, Xichen","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5138631487","display_name":"Aashu Singh","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Singh, Aashu","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5065117717","display_name":"Satya Narayan Shukla","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Shukla, Satya Narayan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5138666291","display_name":"Xiangjun Fan","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Fan, Xiangjun","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5138654969","display_name":"Shlok Kumar Mishra","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Mishra, Shlok Kumar","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5138656039","display_name":"Saining Xie","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Xie, Saining","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":0.7764000296592712,"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":0.7764000296592712,"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/T11714","display_name":"Multimodal Machine Learning Applications","score":0.06989999860525131,"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/T11448","display_name":"Face recognition and analysis","score":0.02449999935925007,"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/representation","display_name":"Representation (politics)","score":0.7099000215530396},{"id":"https://openalex.org/keywords/prior-probability","display_name":"Prior probability","score":0.708299994468689},{"id":"https://openalex.org/keywords/noise-reduction","display_name":"Noise reduction","score":0.559499979019165},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.5297999978065491},{"id":"https://openalex.org/keywords/generative-model","display_name":"Generative model","score":0.5227000117301941},{"id":"https://openalex.org/keywords/generative-grammar","display_name":"Generative grammar","score":0.5},{"id":"https://openalex.org/keywords/space","display_name":"Space (punctuation)","score":0.4968000054359436}],"concepts":[{"id":"https://openalex.org/C2776359362","wikidata":"https://www.wikidata.org/wiki/Q2145286","display_name":"Representation (politics)","level":3,"score":0.7099000215530396},{"id":"https://openalex.org/C177769412","wikidata":"https://www.wikidata.org/wiki/Q278090","display_name":"Prior probability","level":3,"score":0.708299994468689},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.6862000226974487},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.683899998664856},{"id":"https://openalex.org/C163294075","wikidata":"https://www.wikidata.org/wiki/Q581861","display_name":"Noise reduction","level":2,"score":0.559499979019165},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.5297999978065491},{"id":"https://openalex.org/C167966045","wikidata":"https://www.wikidata.org/wiki/Q5532625","display_name":"Generative model","level":3,"score":0.5227000117301941},{"id":"https://openalex.org/C39890363","wikidata":"https://www.wikidata.org/wiki/Q36108","display_name":"Generative grammar","level":2,"score":0.5},{"id":"https://openalex.org/C2778572836","wikidata":"https://www.wikidata.org/wiki/Q380933","display_name":"Space (punctuation)","level":2,"score":0.4968000054359436},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.4101000130176544},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.387800008058548},{"id":"https://openalex.org/C99498987","wikidata":"https://www.wikidata.org/wiki/Q2210247","display_name":"Noise (video)","level":3,"score":0.38100001215934753},{"id":"https://openalex.org/C2779843651","wikidata":"https://www.wikidata.org/wiki/Q7390335","display_name":"SIGNAL (programming language)","level":2,"score":0.37599998712539673},{"id":"https://openalex.org/C165064840","wikidata":"https://www.wikidata.org/wiki/Q1321061","display_name":"Matching (statistics)","level":2,"score":0.3086000084877014},{"id":"https://openalex.org/C126042441","wikidata":"https://www.wikidata.org/wiki/Q1324888","display_name":"Frame (networking)","level":2,"score":0.30320000648498535},{"id":"https://openalex.org/C111335779","wikidata":"https://www.wikidata.org/wiki/Q3454686","display_name":"Reduction (mathematics)","level":2,"score":0.29339998960494995},{"id":"https://openalex.org/C31972630","wikidata":"https://www.wikidata.org/wiki/Q844240","display_name":"Computer vision","level":1,"score":0.2874999940395355},{"id":"https://openalex.org/C59404180","wikidata":"https://www.wikidata.org/wiki/Q17013334","display_name":"Feature learning","level":2,"score":0.2777999937534332},{"id":"https://openalex.org/C36464697","wikidata":"https://www.wikidata.org/wiki/Q451553","display_name":"Visualization","level":2,"score":0.2563000023365021}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2606.14700","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.14700","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":"Preprint"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2606.14700","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.14700","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":"Preprint"},"sustainable_development_goals":[{"display_name":"Industry, innovation and infrastructure","score":0.4515296518802643,"id":"https://metadata.un.org/sdg/9"}],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Large":[0],"language":[1],"models":[2],"(LLMs)":[3],"are":[4,13],"widely":[5],"used":[6],"in":[7,158],"text-to-image":[8],"(T2I)":[9],"systems,":[10],"but":[11],"they":[12],"typically":[14],"limited":[15],"to":[16,66,90,124],"text":[17],"encoding,":[18],"while":[19],"denoising":[20,137],"is":[21,48,64],"handled":[22],"by":[23,56,142],"newly":[24,125],"trained":[25],"generative":[26],"backbones.":[27],"The":[28],"emergence":[29],"of":[30],"representation":[31,83],"autoencoders":[32],"(RAEs)":[33],"shifts":[34],"the":[35,77,98,103],"generation":[36],"target":[37],"toward":[38],"semantically":[39],"structured":[40],"visual":[41,69,138],"representations,":[42,147],"creating":[43],"a":[44,72,81,107],"latent":[45],"space":[46],"that":[47,120,131],"more":[49],"compatible":[50],"with":[51,71],"pretrained":[52,73],"LLM":[53],"priors.":[54],"Inspired":[55],"multimodal":[57],"LLMs":[58],"(MLLMs),":[59],"where":[60],"an":[61],"MLP":[62],"projector":[63],"sufficient":[65],"align":[67],"clean":[68,89],"representations":[70,139],"LLM,":[74],"we":[75],"repurpose":[76],"MLLM":[78,100,156],"itself":[79],"as":[80,102],"noisy":[82,91,146],"encoder,":[84],"extending":[85],"this":[86],"mechanism":[87],"from":[88],"inputs.":[92],"We":[93],"present":[94],"RepFusion,":[95],"which":[96],"uses":[97],"resulting":[99],"outputs":[101],"conditioning":[104,143,157],"signal":[105],"for":[106,136],"diffusion":[108],"transformer.":[109],"In":[110],"controlled":[111],"comparisons":[112],"at":[113],"similar":[114],"inference":[115],"budgets,":[116],"RepFusion":[117],"outperforms":[118],"baselines":[119],"devote":[121],"comparable":[122],"capacity":[123],"initialized":[126],"denoisers.":[127],"These":[128],"results":[129],"demonstrate":[130],"MLLMs":[132],"provide":[133],"strong":[134],"priors":[135],"and":[140],"that,":[141],"on":[144,154],"evolving":[145],"test-time":[148],"compute":[149],"can":[150],"be":[151],"productively":[152],"spent":[153],"repeated":[155],"modern":[159],"T2I":[160],"systems.":[161]},"counts_by_year":[],"updated_date":"2026-07-01T06:00:48.157686","created_date":"2026-06-16T00:00:00"}