{"id":"https://openalex.org/W7170905505","doi":"https://doi.org/10.48550/arxiv.2607.20913","title":"Source-Prior-Driven Selective Adaptation for Efficient Diffusion Model Finetuning","display_name":"Source-Prior-Driven Selective Adaptation for Efficient Diffusion Model Finetuning","publication_year":2026,"publication_date":"2026-07-23","ids":{"openalex":"https://openalex.org/W7170905505","doi":"https://doi.org/10.48550/arxiv.2607.20913"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2607.20913","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2607.20913","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":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.2607.20913","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5128312348","display_name":"Y L Xiong","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Xiong, Yi","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5107927494","display_name":"Y. Cheng","orcid":"https://orcid.org/0000-0002-4907-476X"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Cheng, Yuan-Yuan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5143598835","display_name":"Xiao-Ming Fu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Fu, Xiao-Ming","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":0.8072999715805054,"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":0.8072999715805054,"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/T11206","display_name":"Model Reduction and Neural Networks","score":0.07320000231266022,"subfield":{"id":"https://openalex.org/subfields/3109","display_name":"Statistical and Nonlinear Physics"},"field":{"id":"https://openalex.org/fields/31","display_name":"Physics and Astronomy"},"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":0.02979999966919422,"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/generative-grammar","display_name":"Generative grammar","score":0.6743000149726868},{"id":"https://openalex.org/keywords/key","display_name":"Key (lock)","score":0.6205999851226807},{"id":"https://openalex.org/keywords/adaptation","display_name":"Adaptation (eye)","score":0.574400007724762},{"id":"https://openalex.org/keywords/construct","display_name":"Construct (python library)","score":0.527999997138977},{"id":"https://openalex.org/keywords/generative-model","display_name":"Generative model","score":0.43369999527931213},{"id":"https://openalex.org/keywords/diffusion","display_name":"Diffusion","score":0.3582000136375427},{"id":"https://openalex.org/keywords/exploit","display_name":"Exploit","score":0.29109999537467957}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7926999926567078},{"id":"https://openalex.org/C39890363","wikidata":"https://www.wikidata.org/wiki/Q36108","display_name":"Generative grammar","level":2,"score":0.6743000149726868},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.6205999851226807},{"id":"https://openalex.org/C139807058","wikidata":"https://www.wikidata.org/wiki/Q352374","display_name":"Adaptation (eye)","level":2,"score":0.574400007724762},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5321000218391418},{"id":"https://openalex.org/C2780801425","wikidata":"https://www.wikidata.org/wiki/Q5164392","display_name":"Construct (python library)","level":2,"score":0.527999997138977},{"id":"https://openalex.org/C167966045","wikidata":"https://www.wikidata.org/wiki/Q5532625","display_name":"Generative model","level":3,"score":0.43369999527931213},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.40630000829696655},{"id":"https://openalex.org/C69357855","wikidata":"https://www.wikidata.org/wiki/Q163214","display_name":"Diffusion","level":2,"score":0.3582000136375427},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.3043000102043152},{"id":"https://openalex.org/C165696696","wikidata":"https://www.wikidata.org/wiki/Q11287","display_name":"Exploit","level":2,"score":0.29109999537467957},{"id":"https://openalex.org/C12426560","wikidata":"https://www.wikidata.org/wiki/Q189569","display_name":"Basis (linear algebra)","level":2,"score":0.28110000491142273},{"id":"https://openalex.org/C12713177","wikidata":"https://www.wikidata.org/wiki/Q1900281","display_name":"Perspective (graphical)","level":2,"score":0.27799999713897705},{"id":"https://openalex.org/C61797465","wikidata":"https://www.wikidata.org/wiki/Q1188986","display_name":"Term (time)","level":2,"score":0.2676999866962433},{"id":"https://openalex.org/C135798126","wikidata":"https://www.wikidata.org/wiki/Q2167279","display_name":"Top-down and bottom-up design","level":2,"score":0.2590999901294708},{"id":"https://openalex.org/C165064840","wikidata":"https://www.wikidata.org/wiki/Q1321061","display_name":"Matching (statistics)","level":2,"score":0.2522999942302704},{"id":"https://openalex.org/C2982832238","wikidata":"https://www.wikidata.org/wiki/Q5531640","display_name":"General purpose","level":2,"score":0.2515999972820282}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2607.20913","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2607.20913","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:10.48550/arxiv.2607.20913","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2607.20913","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":false,"raw_source_name":null,"raw_type":"Preprint"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Fine-tuning":[0],"large":[1],"diffusion":[2,49],"models":[3],"for":[4,114,123],"new":[5],"domains":[6],"or":[7],"styles":[8],"involves":[9],"a":[10,40,52,80,105,133],"trade-off:":[11],"improving":[12],"target-specific":[13],"generation":[14],"often":[15],"degrades":[16],"the":[17,63,85,124],"pretrained":[18,73],"model's":[19,86],"broad":[20],"generative":[21,67,88],"capability.":[22],"Existing":[23],"full":[24],"and":[25,75,95,117],"parameter-efficient":[26],"fine-tuning":[27],"methods":[28],"typically":[29],"handle":[30],"this":[31,36],"trade-off":[32,136],"only":[33],"implicitly.":[34],"In":[35],"work,":[37],"we":[38,102],"propose":[39],"novel":[41],"source-prior-driven":[42],"selective":[43],"adaptation":[44],"method":[45,56,131],"to":[46,108],"efficiently":[47],"fine-tune":[48],"models,":[50],"achieving":[51],"favorable":[53],"trade-off.":[54],"Our":[55],"relies":[57],"on":[58,84],"two":[59],"key":[60],"observations:":[61],"(1)":[62],"loss":[64],"of":[65],"general":[66,87],"capability":[68,89],"is":[69],"highly":[70],"inconsistent":[71,92],"across":[72,93],"parameters,":[74],"(2)":[76],"parameters":[77,111],"that":[78,129],"have":[79],"relatively":[81],"small":[82],"impact":[83],"remain":[90],"structurally":[91],"layers":[94],"parameter":[96],"types.":[97],"Motivated":[98],"by":[99],"these":[100],"observations,":[101],"first":[103],"learn":[104],"static":[106],"mask":[107],"explicitly":[109],"identify":[110],"better":[112,134],"suited":[113],"downstream":[115],"adaptation,":[116],"then":[118],"construct":[119],"structured":[120],"update":[121],"strategies":[122],"selected":[125],"subset.":[126],"Experiments":[127],"show":[128],"our":[130],"achieves":[132],"adaptation-retention":[135],"than":[137],"existing":[138],"strong":[139],"baselines.":[140]},"counts_by_year":[],"updated_date":"2026-07-28T07:46:37.118299","created_date":"2026-07-25T00:00:00"}