{"id":"https://openalex.org/W7136569424","doi":"https://doi.org/10.48550/arxiv.2603.12433","title":"Revisiting Model Stitching In the Foundation Model Era","display_name":"Revisiting Model Stitching In the Foundation Model Era","publication_year":2026,"publication_date":"2026-03-12","ids":{"openalex":"https://openalex.org/W7136569424","doi":"https://doi.org/10.48550/arxiv.2603.12433"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2603.12433","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2603.12433","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":"article"},"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.48550/arxiv.2603.12433","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5066572795","display_name":"Zheda Mai","orcid":"https://orcid.org/0000-0003-3703-563X"},"institutions":[],"countries":[],"is_corresponding":true,"raw_author_name":"Mai, Zheda","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5129466379","display_name":"Ke Zhang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhang, Ke","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5014665083","display_name":"Fu-En Wang","orcid":"https://orcid.org/0000-0002-0992-1899"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wang, Fu-En","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5129637884","display_name":"Zixiao Ken Wang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wang, Zixiao Ken","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5070952449","display_name":"Albert Y. C. Chen","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Chen, Albert Y. C.","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5129530902","display_name":"Lu Xia","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Xia, Lu","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5129609744","display_name":"Min Sun","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Sun, Min","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5129531752","display_name":"Wei-Lun Chao","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Chao, Wei-Lun","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5105792706","display_name":"Cheng-Hao Kuo","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Kuo, Cheng-Hao","raw_affiliation_strings":[],"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":9,"corresponding_author_ids":["https://openalex.org/A5066572795"],"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/T11714","display_name":"Multimodal Machine Learning Applications","score":0.6345000267028809,"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/T11714","display_name":"Multimodal Machine Learning Applications","score":0.6345000267028809,"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":0.05860000103712082,"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/T11094","display_name":"Face Recognition and Perception","score":0.05689999833703041,"subfield":{"id":"https://openalex.org/subfields/2805","display_name":"Cognitive Neuroscience"},"field":{"id":"https://openalex.org/fields/28","display_name":"Neuroscience"},"domain":{"id":"https://openalex.org/domains/1","display_name":"Life Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/image-stitching","display_name":"Image stitching","score":0.9449999928474426},{"id":"https://openalex.org/keywords/leverage","display_name":"Leverage (statistics)","score":0.6607999801635742},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.4632999897003174},{"id":"https://openalex.org/keywords/overhead","display_name":"Overhead (engineering)","score":0.4413999915122986},{"id":"https://openalex.org/keywords/layer","display_name":"Layer (electronics)","score":0.3905999958515167},{"id":"https://openalex.org/keywords/foundation","display_name":"Foundation (evidence)","score":0.3538999855518341},{"id":"https://openalex.org/keywords/task","display_name":"Task (project management)","score":0.34119999408721924}],"concepts":[{"id":"https://openalex.org/C29081049","wikidata":"https://www.wikidata.org/wiki/Q1364242","display_name":"Image stitching","level":2,"score":0.9449999928474426},{"id":"https://openalex.org/C153083717","wikidata":"https://www.wikidata.org/wiki/Q6535263","display_name":"Leverage (statistics)","level":2,"score":0.6607999801635742},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.6453999876976013},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6015999913215637},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.4632999897003174},{"id":"https://openalex.org/C2779960059","wikidata":"https://www.wikidata.org/wiki/Q7113681","display_name":"Overhead (engineering)","level":2,"score":0.4413999915122986},{"id":"https://openalex.org/C2779227376","wikidata":"https://www.wikidata.org/wiki/Q6505497","display_name":"Layer (electronics)","level":2,"score":0.3905999958515167},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.3873000144958496},{"id":"https://openalex.org/C2780966255","wikidata":"https://www.wikidata.org/wiki/Q5474306","display_name":"Foundation (evidence)","level":2,"score":0.3538999855518341},{"id":"https://openalex.org/C2780451532","wikidata":"https://www.wikidata.org/wiki/Q759676","display_name":"Task (project management)","level":2,"score":0.34119999408721924},{"id":"https://openalex.org/C199639397","wikidata":"https://www.wikidata.org/wiki/Q1788588","display_name":"Engineering drawing","level":1,"score":0.3292999863624573},{"id":"https://openalex.org/C108583219","wikidata":"https://www.wikidata.org/wiki/Q197536","display_name":"Deep learning","level":2,"score":0.3273000121116638},{"id":"https://openalex.org/C31972630","wikidata":"https://www.wikidata.org/wiki/Q844240","display_name":"Computer vision","level":1,"score":0.3237999975681305},{"id":"https://openalex.org/C28719098","wikidata":"https://www.wikidata.org/wiki/Q44946","display_name":"Point (geometry)","level":2,"score":0.32249999046325684},{"id":"https://openalex.org/C113174947","wikidata":"https://www.wikidata.org/wiki/Q2859736","display_name":"Tree (set theory)","level":2,"score":0.31279999017715454},{"id":"https://openalex.org/C2780586882","wikidata":"https://www.wikidata.org/wiki/Q7520643","display_name":"Simple (philosophy)","level":2,"score":0.2962999939918518},{"id":"https://openalex.org/C63479239","wikidata":"https://www.wikidata.org/wiki/Q7353546","display_name":"Robustness (evolution)","level":3,"score":0.28940001130104065},{"id":"https://openalex.org/C127413603","wikidata":"https://www.wikidata.org/wiki/Q11023","display_name":"Engineering","level":0,"score":0.2800999879837036},{"id":"https://openalex.org/C77618280","wikidata":"https://www.wikidata.org/wiki/Q1155772","display_name":"Scheme (mathematics)","level":2,"score":0.272599995136261},{"id":"https://openalex.org/C22367795","wikidata":"https://www.wikidata.org/wiki/Q7625208","display_name":"Structured prediction","level":2,"score":0.27129998803138733},{"id":"https://openalex.org/C168167062","wikidata":"https://www.wikidata.org/wiki/Q1117970","display_name":"Component (thermodynamics)","level":2,"score":0.26510000228881836}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2603.12433","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2603.12433","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":"article"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2603.12433","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2603.12433","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":"article"},"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":{"Model":[0],"stitching,":[1],"connecting":[2],"early":[3,183],"layers":[4,11,184],"of":[5,12,25],"one":[6],"model":[7,153,158],"(source)":[8],"to":[9,118,215],"later":[10,190],"another":[13],"(target)":[14],"via":[15],"a":[16,23,77,128,161,193,212,216],"light":[17],"stitch":[18,82,84,109,124,149,167],"layer,":[19,137],"has":[20],"served":[21],"as":[22],"probe":[24,214],"representational":[26],"compatibility.":[27],"Prior":[28],"work":[29],"finds":[30],"that":[31,56,102,200],"models":[32],"trained":[33],"on":[34,170],"the":[35,81,104,108,113,133,151,166,176],"same":[36],"dataset":[37],"remain":[38],"stitchable":[39,142],"(negligible":[40],"accuracy":[41],"drop)":[42],"despite":[43],"different":[44],"initializations":[45],"or":[46,111,230],"objectives.":[47],"We":[48,75],"revisit":[49],"stitching":[50,210],"for":[51,197,219],"Vision":[52],"Foundation":[53],"Models":[54],"(VFMs)":[55],"vary":[57],"in":[58],"objectives,":[59],"data,":[60],"and":[61,69,89,224],"modality":[62],"mix":[63],"(e.g.,":[64],"CLIP,":[65],"DINOv2,":[66],"SigLIP":[67],"2)":[68],"ask:":[70],"Are":[71],"heterogeneous":[72,138],"VFMs":[73,139,186],"stitchable?":[74],"introduce":[76],"systematic":[78],"protocol":[79],"spanning":[80],"points,":[83,150],"layer":[85,97],"families,":[86],"training":[87,98],"losses,":[88],"downstream":[90],"tasks.":[91,145],"Three":[92],"findings":[93],"emerge.":[94],"(1)":[95],"Stitch":[96,178],"matters:":[99],"conventional":[100],"approaches":[101],"match":[103],"intermediate":[105],"features":[106],"at":[107,122,132,159],"point":[110],"optimize":[112],"task":[114],"loss":[115,131],"end-to-end":[116],"struggle":[117],"retain":[119],"accuracy,":[120],"especially":[121],"shallow":[123],"points.":[125],"(2)":[126],"With":[127],"simple":[129],"feature-matching":[130],"target":[134],"model's":[135],"penultimate":[136],"become":[140],"reliably":[141],"across":[143,185],"vision":[144],"(3)":[146],"For":[147],"deep":[148],"stitched":[152],"can":[154],"surpass":[155],"either":[156],"constituent":[157],"only":[160],"small":[162],"inference":[163],"overhead":[164],"(for":[165],"layer).":[168],"Building":[169],"these":[171],"findings,":[172],"we":[173],"further":[174],"propose":[175],"VFM":[177,222],"Tree":[179],"(VST),":[180],"which":[181],"shares":[182],"while":[187],"retaining":[188],"their":[189,227],"layers,":[191],"yielding":[192],"controllable":[194],"accuracy-latency":[195],"trade-off":[196],"multimodal":[198],"LLMs":[199],"often":[201],"leverage":[202],"multiple":[203],"VFMs.":[204],"Taken":[205],"together,":[206],"our":[207],"study":[208],"elevates":[209],"from":[211],"diagnostic":[213],"practical":[217],"recipe":[218],"integrating":[220],"complementary":[221],"strengths":[223],"pinpointing":[225],"where":[226],"representations":[228],"align":[229],"diverge.":[231]},"counts_by_year":[],"updated_date":"2026-03-17T07:05:13.627479","created_date":"2026-03-17T00:00:00"}