{"id":"https://openalex.org/W7155217034","doi":"https://doi.org/10.48550/arxiv.2604.19254","title":"ShadowPEFT: Shadow Network for Parameter-Efficient Fine-Tuning","display_name":"ShadowPEFT: Shadow Network for Parameter-Efficient Fine-Tuning","publication_year":2026,"publication_date":"2026-04-21","ids":{"openalex":"https://openalex.org/W7155217034","doi":"https://doi.org/10.48550/arxiv.2604.19254"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2604.19254","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.19254","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":"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.2604.19254","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5134274656","display_name":"Xianming Li","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Li, Xianming","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5134349900","display_name":"Zongxi Li","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Li, Zongxi","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5134320355","display_name":"Tsz-fung Andrew Lee","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Lee, Tsz-fung Andrew","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100336945","display_name":"Jing Li","orcid":"https://orcid.org/0000-0002-5330-2489"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Li, Jing","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5134278732","display_name":"Haoran Xie","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Xie, Haoran","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5134246001","display_name":"Qing Li","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Li, Qing","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/T10028","display_name":"Topic Modeling","score":0.21050000190734863,"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/T10028","display_name":"Topic Modeling","score":0.21050000190734863,"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/T11307","display_name":"Domain Adaptation and Few-Shot Learning","score":0.19840000569820404,"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/T10775","display_name":"Generative Adversarial Networks and Image Synthesis","score":0.08980000019073486,"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/shadow","display_name":"Shadow (psychology)","score":0.6644999980926514},{"id":"https://openalex.org/keywords/adaptation","display_name":"Adaptation (eye)","score":0.6144000291824341},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.5623999834060669},{"id":"https://openalex.org/keywords/set","display_name":"Set (abstract data type)","score":0.5016999840736389},{"id":"https://openalex.org/keywords/enhanced-data-rates-for-gsm-evolution","display_name":"Enhanced Data Rates for GSM Evolution","score":0.40220001339912415},{"id":"https://openalex.org/keywords/shadow-mask","display_name":"Shadow mask","score":0.3889999985694885}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7670999765396118},{"id":"https://openalex.org/C117797892","wikidata":"https://www.wikidata.org/wiki/Q286363","display_name":"Shadow (psychology)","level":2,"score":0.6644999980926514},{"id":"https://openalex.org/C139807058","wikidata":"https://www.wikidata.org/wiki/Q352374","display_name":"Adaptation (eye)","level":2,"score":0.6144000291824341},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.5623999834060669},{"id":"https://openalex.org/C177264268","wikidata":"https://www.wikidata.org/wiki/Q1514741","display_name":"Set (abstract data type)","level":2,"score":0.5016999840736389},{"id":"https://openalex.org/C120314980","wikidata":"https://www.wikidata.org/wiki/Q180634","display_name":"Distributed computing","level":1,"score":0.476500004529953},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.45980000495910645},{"id":"https://openalex.org/C162307627","wikidata":"https://www.wikidata.org/wiki/Q204833","display_name":"Enhanced Data Rates for GSM Evolution","level":2,"score":0.40220001339912415},{"id":"https://openalex.org/C31972630","wikidata":"https://www.wikidata.org/wiki/Q844240","display_name":"Computer vision","level":1,"score":0.39100000262260437},{"id":"https://openalex.org/C2776326872","wikidata":"https://www.wikidata.org/wiki/Q1192335","display_name":"Shadow mask","level":2,"score":0.3889999985694885},{"id":"https://openalex.org/C79403827","wikidata":"https://www.wikidata.org/wiki/Q3988","display_name":"Real-time computing","level":1,"score":0.3472000062465668},{"id":"https://openalex.org/C160633673","wikidata":"https://www.wikidata.org/wiki/Q355198","display_name":"Pixel","level":2,"score":0.34700000286102295},{"id":"https://openalex.org/C66322947","wikidata":"https://www.wikidata.org/wiki/Q11658","display_name":"Transformer","level":3,"score":0.34450000524520874},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.3239000141620636},{"id":"https://openalex.org/C116544410","wikidata":"https://www.wikidata.org/wiki/Q1478122","display_name":"Shadow mapping","level":2,"score":0.3231000006198883},{"id":"https://openalex.org/C48103436","wikidata":"https://www.wikidata.org/wiki/Q599031","display_name":"State (computer science)","level":2,"score":0.31529998779296875},{"id":"https://openalex.org/C36464697","wikidata":"https://www.wikidata.org/wiki/Q451553","display_name":"Visualization","level":2,"score":0.29269999265670776},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.2653999924659729},{"id":"https://openalex.org/C80444323","wikidata":"https://www.wikidata.org/wiki/Q2878974","display_name":"Theoretical computer science","level":1,"score":0.25589999556541443}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2604.19254","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.19254","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":"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.2604.19254","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.19254","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":"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":[],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Parameter-efficient":[0],"fine-tuning":[1,9],"(PEFT)":[2],"reduces":[3],"the":[4,26,106,112],"training":[5,16],"cost":[6],"of":[7,21,53],"full-parameter":[8],"for":[10,86],"large":[11],"language":[12],"models":[13],"(LLMs)":[14],"by":[15,39],"only":[17],"a":[18,50,58,68,78,100,126,172],"small":[19],"set":[20],"task-specific":[22],"parameters":[23],"while":[24],"freezing":[25],"pretrained":[27],"backbone.":[28],"However,":[29],"existing":[30],"approaches,":[31],"such":[32],"as":[33],"Low-Rank":[34],"Adaptation":[35],"(LoRA),":[36],"achieve":[37],"adaptation":[38,94,170],"inserting":[40],"independent":[41],"low-rank":[42,179],"perturbations":[43,98],"directly":[44],"to":[45,99,177],"individual":[46],"weights,":[47],"resulting":[48],"in":[49,125],"local":[51],"parameterization":[52],"adaptation.":[54],"We":[55],"propose":[56],"ShadowPEFT,":[57],"centralized":[59,168],"PEFT":[60],"framework":[61],"that":[62,140,167],"instead":[63],"performs":[64],"layer-level":[65],"refinement":[66,103],"through":[67],"depth-shared":[69],"shadow":[70,80,107,155],"module.":[71],"At":[72],"each":[73],"transformer":[74],"layer,":[75],"ShadowPEFT":[76,141],"maintains":[77],"parallel":[79],"state":[81],"and":[82,122,136,146,163,174],"evolves":[83],"it":[84,114],"repeatedly":[85],"progressively":[87],"richer":[88],"hidden":[89],"states.":[90],"This":[91],"design":[92],"shifts":[93],"from":[95,111],"distributed":[96],"weight-space":[97],"shared":[101],"layer-space":[102,169],"process.":[104],"Since":[105],"module":[108],"is":[109,171],"decoupled":[110],"backbone,":[113],"can":[115],"be":[116],"reused":[117],"across":[118],"depth,":[119],"independently":[120],"pretrained,":[121],"optionally":[123],"deployed":[124],"detached":[127],"mode,":[128],"benefiting":[129],"edge":[130],"computing":[131],"scenarios.":[132],"Experiments":[133],"on":[134,154],"generation":[135],"understanding":[137],"benchmarks":[138],"show":[139],"matches":[142],"or":[143],"outperforms":[144],"LoRA":[145],"DoRA":[147],"under":[148],"comparable":[149],"trainable-parameter":[150],"budgets.":[151],"Additional":[152],"analyses":[153],"pretraining,":[156],"cross-dataset":[157],"transfer,":[158],"parameter":[159],"scaling,":[160],"inference":[161],"latency,":[162],"system-level":[164],"evaluation":[165],"suggest":[166],"competitive":[173],"flexible":[175],"alternative":[176],"conventional":[178],"PEFT.":[180]},"counts_by_year":[],"updated_date":"2026-07-28T07:46:37.118299","created_date":"2026-04-23T00:00:00"}