{"id":"https://openalex.org/W7130378192","doi":"https://doi.org/10.48550/arxiv.2602.15539","title":"Dynamic Training-Free Fusion of Subject and Style LoRAs","display_name":"Dynamic Training-Free Fusion of Subject and Style LoRAs","publication_year":2026,"publication_date":"2026-02-17","ids":{"openalex":"https://openalex.org/W7130378192","doi":"https://doi.org/10.48550/arxiv.2602.15539"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2602.15539","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2602.15539","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":"article"},"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.48550/arxiv.2602.15539","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5126325256","display_name":"Qinglong Cao","orcid":null},"institutions":[],"countries":[],"is_corresponding":true,"raw_author_name":"Cao, Qinglong","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5126303780","display_name":"Yuntian Chen","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Chen, Yuntian","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5126309822","display_name":"Chao Ma","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ma, Chao","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5126302922","display_name":"Xiaokang Yang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yang, Xiaokang","raw_affiliation_strings":[],"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":4,"corresponding_author_ids":["https://openalex.org/A5126325256"],"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":0.9556000232696533,"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":0.9556000232696533,"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":0.011699999682605267,"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/T11307","display_name":"Domain Adaptation and Few-Shot Learning","score":0.003800000064074993,"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/heuristics","display_name":"Heuristics","score":0.5166000127792358},{"id":"https://openalex.org/keywords/selection","display_name":"Selection (genetic algorithm)","score":0.474700003862381},{"id":"https://openalex.org/keywords/randomness","display_name":"Randomness","score":0.4406999945640564},{"id":"https://openalex.org/keywords/feature","display_name":"Feature (linguistics)","score":0.4343999922275543},{"id":"https://openalex.org/keywords/benchmark","display_name":"Benchmark (surveying)","score":0.42289999127388},{"id":"https://openalex.org/keywords/divergence","display_name":"Divergence (linguistics)","score":0.40450000762939453},{"id":"https://openalex.org/keywords/fusion","display_name":"Fusion","score":0.3840000033378601},{"id":"https://openalex.org/keywords/ranking","display_name":"Ranking (information retrieval)","score":0.3734000027179718}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7311000227928162},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.579200029373169},{"id":"https://openalex.org/C127705205","wikidata":"https://www.wikidata.org/wiki/Q5748245","display_name":"Heuristics","level":2,"score":0.5166000127792358},{"id":"https://openalex.org/C81917197","wikidata":"https://www.wikidata.org/wiki/Q628760","display_name":"Selection (genetic algorithm)","level":2,"score":0.474700003862381},{"id":"https://openalex.org/C125112378","wikidata":"https://www.wikidata.org/wiki/Q176640","display_name":"Randomness","level":2,"score":0.4406999945640564},{"id":"https://openalex.org/C2776401178","wikidata":"https://www.wikidata.org/wiki/Q12050496","display_name":"Feature (linguistics)","level":2,"score":0.4343999922275543},{"id":"https://openalex.org/C185798385","wikidata":"https://www.wikidata.org/wiki/Q1161707","display_name":"Benchmark (surveying)","level":2,"score":0.42289999127388},{"id":"https://openalex.org/C207390915","wikidata":"https://www.wikidata.org/wiki/Q1230525","display_name":"Divergence (linguistics)","level":2,"score":0.40450000762939453},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.39579999446868896},{"id":"https://openalex.org/C158525013","wikidata":"https://www.wikidata.org/wiki/Q2593739","display_name":"Fusion","level":2,"score":0.3840000033378601},{"id":"https://openalex.org/C189430467","wikidata":"https://www.wikidata.org/wiki/Q7293293","display_name":"Ranking (information retrieval)","level":2,"score":0.3734000027179718},{"id":"https://openalex.org/C148483581","wikidata":"https://www.wikidata.org/wiki/Q446488","display_name":"Feature selection","level":2,"score":0.36579999327659607},{"id":"https://openalex.org/C2777855551","wikidata":"https://www.wikidata.org/wiki/Q12310021","display_name":"Subject (documents)","level":2,"score":0.35179999470710754},{"id":"https://openalex.org/C13662910","wikidata":"https://www.wikidata.org/wiki/Q193139","display_name":"Trajectory","level":2,"score":0.34060001373291016},{"id":"https://openalex.org/C141353440","wikidata":"https://www.wikidata.org/wiki/Q182221","display_name":"Fuse (electrical)","level":2,"score":0.3246000111103058},{"id":"https://openalex.org/C42058472","wikidata":"https://www.wikidata.org/wiki/Q810214","display_name":"Base (topology)","level":2,"score":0.3240000009536743},{"id":"https://openalex.org/C146834321","wikidata":"https://www.wikidata.org/wiki/Q2979672","display_name":"Closure (psychology)","level":2,"score":0.29280000925064087},{"id":"https://openalex.org/C33954974","wikidata":"https://www.wikidata.org/wiki/Q486494","display_name":"Sensor fusion","level":2,"score":0.2727000117301941},{"id":"https://openalex.org/C28719098","wikidata":"https://www.wikidata.org/wiki/Q44946","display_name":"Point (geometry)","level":2,"score":0.2632000148296356},{"id":"https://openalex.org/C7149132","wikidata":"https://www.wikidata.org/wiki/Q1377840","display_name":"Forgetting","level":2,"score":0.25360000133514404},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.25279998779296875}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2602.15539","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2602.15539","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":"article"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2602.15539","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2602.15539","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":"article"},"sustainable_development_goals":[{"display_name":"Quality Education","score":0.7105921506881714,"id":"https://metadata.un.org/sdg/4"}],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Recent":[0],"studies":[1],"have":[2],"explored":[3],"the":[4,40,58,62,72,76,93,100,107,142],"combination":[5],"of":[6,33,42],"multiple":[7],"LoRAs":[8],"to":[9],"simultaneously":[10],"generate":[11],"user-specified":[12],"subjects":[13],"and":[14,38,81,86,90,122,128,138,174],"styles.":[15],"However,":[16],"most":[17,94],"existing":[18],"approaches":[19],"fuse":[20],"LoRA":[21,169],"weights":[22,96],"using":[23],"static":[24],"statistical":[25],"heuristics":[26],"that":[27,55,163],"deviate":[28],"from":[29,116],"LoRA's":[30],"original":[31,79],"purpose":[32],"learning":[34],"adaptive":[35],"feature":[36],"adjustments":[37],"ignore":[39],"randomness":[41],"sampled":[43],"inputs.":[44],"To":[45],"address":[46],"this,":[47],"we":[48,69,104],"propose":[49],"a":[50],"dynamic":[51],"training-free":[52],"fusion":[53,170],"framework":[54],"operates":[56],"throughout":[57],"generation":[59,108],"process.":[60],"During":[61],"forward":[63],"pass,":[64],"at":[65],"each":[66],"LoRA-applied":[67],"layer,":[68],"dynamically":[70,111,148],"compute":[71],"KL":[73],"divergence":[74],"between":[75],"base":[77],"model's":[78],"features":[80],"those":[82],"produced":[83],"by":[84,110],"subject":[85],"style":[87],"LoRAs,":[88],"respectively,":[89],"adaptively":[91],"select":[92],"appropriate":[95],"for":[97],"fusion.":[98],"In":[99],"reverse":[101],"denoising":[102],"stage,":[103],"further":[105],"refine":[106],"trajectory":[109],"applying":[112],"gradient-based":[113],"corrections":[114],"derived":[115],"objective":[117],"metrics":[118],"such":[119],"as":[120],"CLIP":[121],"DINO":[123],"scores,":[124],"providing":[125],"continuous":[126],"semantic":[127],"stylistic":[129],"guidance.":[130],"By":[131],"integrating":[132],"these":[133],"two":[134],"complementary":[135],"mechanisms-feature-level":[136],"selection":[137],"metric-guided":[139],"latent":[140],"adjustment-across":[141],"entire":[143],"diffusion":[144],"timeline,":[145],"our":[146,164],"method":[147],"achieves":[149],"coherent":[150],"subject-style":[151,160],"synthesis":[152],"without":[153],"any":[154],"retraining.":[155],"Extensive":[156],"experiments":[157],"across":[158],"diverse":[159],"combinations":[161],"demonstrate":[162],"approach":[165],"consistently":[166],"outperforms":[167],"state-of-the-art":[168],"methods":[171],"both":[172],"qualitatively":[173],"quantitatively.":[175]},"counts_by_year":[],"updated_date":"2026-02-19T06:31:58.851227","created_date":"2026-02-19T00:00:00"}