{"id":"https://openalex.org/W7128807805","doi":"https://doi.org/10.48550/arxiv.2602.11636","title":"ScalSelect: Scalable Training-Free Multimodal Data Selection for Efficient Visual Instruction Tuning","display_name":"ScalSelect: Scalable Training-Free Multimodal Data Selection for Efficient Visual Instruction Tuning","publication_year":2026,"publication_date":"2026-02-12","ids":{"openalex":"https://openalex.org/W7128807805","doi":"https://doi.org/10.48550/arxiv.2602.11636"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2602.11636","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2602.11636","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.2602.11636","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5069583018","display_name":"Changti Wu","orcid":"https://orcid.org/0009-0009-9448-6657"},"institutions":[],"countries":[],"is_corresponding":true,"raw_author_name":"Wu, Changti","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5125952199","display_name":"Jiahuai Mao","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Mao, Jiahuai","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5123509874","display_name":"Yuzhuo Miao","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Miao, Yuzhuo","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5112968072","display_name":"Shijie Lian","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Lian, Shijie","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5125934341","display_name":"Bin Yu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yu, Bin","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5125889164","display_name":"Xiaopeng Lin","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Lin, Xiaopeng","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5125930867","display_name":"Cong Huang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Huang, Cong","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5125938047","display_name":"Lei Zhang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhang, Lei","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5125979606","display_name":"Kai Chen","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Chen, Kai","raw_affiliation_strings":[],"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":9,"corresponding_author_ids":["https://openalex.org/A5069583018"],"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":0.9664999842643738,"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":0.9664999842643738,"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/T10036","display_name":"Advanced Neural Network Applications","score":0.005499999970197678,"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":0.005200000014156103,"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/pairwise-comparison","display_name":"Pairwise comparison","score":0.7527999877929688},{"id":"https://openalex.org/keywords/scalability","display_name":"Scalability","score":0.7064999938011169},{"id":"https://openalex.org/keywords/subspace-topology","display_name":"Subspace topology","score":0.6000999808311462},{"id":"https://openalex.org/keywords/redundancy","display_name":"Redundancy (engineering)","score":0.5564000010490417},{"id":"https://openalex.org/keywords/selection","display_name":"Selection (genetic algorithm)","score":0.5527999997138977},{"id":"https://openalex.org/keywords/training-set","display_name":"Training set","score":0.45570001006126404},{"id":"https://openalex.org/keywords/data-redundancy","display_name":"Data redundancy","score":0.40939998626708984},{"id":"https://openalex.org/keywords/representation","display_name":"Representation (politics)","score":0.40299999713897705},{"id":"https://openalex.org/keywords/model-selection","display_name":"Model selection","score":0.3930000066757202}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8289999961853027},{"id":"https://openalex.org/C184898388","wikidata":"https://www.wikidata.org/wiki/Q1435712","display_name":"Pairwise comparison","level":2,"score":0.7527999877929688},{"id":"https://openalex.org/C48044578","wikidata":"https://www.wikidata.org/wiki/Q727490","display_name":"Scalability","level":2,"score":0.7064999938011169},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.6728000044822693},{"id":"https://openalex.org/C32834561","wikidata":"https://www.wikidata.org/wiki/Q660730","display_name":"Subspace topology","level":2,"score":0.6000999808311462},{"id":"https://openalex.org/C152124472","wikidata":"https://www.wikidata.org/wiki/Q1204361","display_name":"Redundancy (engineering)","level":2,"score":0.5564000010490417},{"id":"https://openalex.org/C81917197","wikidata":"https://www.wikidata.org/wiki/Q628760","display_name":"Selection (genetic algorithm)","level":2,"score":0.5527999997138977},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5153999924659729},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.4934000074863434},{"id":"https://openalex.org/C51632099","wikidata":"https://www.wikidata.org/wiki/Q3985153","display_name":"Training set","level":2,"score":0.45570001006126404},{"id":"https://openalex.org/C7545210","wikidata":"https://www.wikidata.org/wiki/Q838123","display_name":"Data redundancy","level":2,"score":0.40939998626708984},{"id":"https://openalex.org/C2776359362","wikidata":"https://www.wikidata.org/wiki/Q2145286","display_name":"Representation (politics)","level":3,"score":0.40299999713897705},{"id":"https://openalex.org/C93959086","wikidata":"https://www.wikidata.org/wiki/Q6888345","display_name":"Model selection","level":2,"score":0.3930000066757202},{"id":"https://openalex.org/C188198153","wikidata":"https://www.wikidata.org/wiki/Q1613840","display_name":"Limiting","level":2,"score":0.37380000948905945},{"id":"https://openalex.org/C103278499","wikidata":"https://www.wikidata.org/wiki/Q254465","display_name":"Similarity (geometry)","level":3,"score":0.351500004529953},{"id":"https://openalex.org/C36464697","wikidata":"https://www.wikidata.org/wiki/Q451553","display_name":"Visualization","level":2,"score":0.34279999136924744},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.3292999863624573},{"id":"https://openalex.org/C148483581","wikidata":"https://www.wikidata.org/wiki/Q446488","display_name":"Feature selection","level":2,"score":0.3264999985694885},{"id":"https://openalex.org/C129844170","wikidata":"https://www.wikidata.org/wiki/Q41299","display_name":"Quadratic equation","level":2,"score":0.31130000948905945},{"id":"https://openalex.org/C116409475","wikidata":"https://www.wikidata.org/wiki/Q1385056","display_name":"External Data Representation","level":2,"score":0.30970001220703125},{"id":"https://openalex.org/C45374587","wikidata":"https://www.wikidata.org/wiki/Q12525525","display_name":"Computation","level":2,"score":0.3037000000476837},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.29589998722076416},{"id":"https://openalex.org/C67186912","wikidata":"https://www.wikidata.org/wiki/Q367664","display_name":"Data modeling","level":2,"score":0.2921999990940094},{"id":"https://openalex.org/C21080849","wikidata":"https://www.wikidata.org/wiki/Q13611879","display_name":"Data point","level":2,"score":0.2759999930858612},{"id":"https://openalex.org/C2776760102","wikidata":"https://www.wikidata.org/wiki/Q5139990","display_name":"Code (set theory)","level":3,"score":0.2754000127315521},{"id":"https://openalex.org/C2777211547","wikidata":"https://www.wikidata.org/wiki/Q17141490","display_name":"Training (meteorology)","level":2,"score":0.26649999618530273},{"id":"https://openalex.org/C59404180","wikidata":"https://www.wikidata.org/wiki/Q17013334","display_name":"Feature learning","level":2,"score":0.2606000006198883},{"id":"https://openalex.org/C2778915421","wikidata":"https://www.wikidata.org/wiki/Q3643177","display_name":"Performance improvement","level":2,"score":0.2563000023365021},{"id":"https://openalex.org/C2778755073","wikidata":"https://www.wikidata.org/wiki/Q10858537","display_name":"Scale (ratio)","level":2,"score":0.25200000405311584}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2602.11636","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2602.11636","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.2602.11636","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2602.11636","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":[],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Large-scale":[0],"Visual":[1],"Instruction":[2],"Tuning":[3],"(VIT)":[4],"has":[5],"become":[6],"a":[7,92],"key":[8],"paradigm":[9],"for":[10,43,55,112],"advancing":[11],"the":[12,25,37,41,105,110,133,147,151,178,183,190],"performance":[13,179],"of":[14,107,150,177,180,189],"vision-language":[15],"models":[16,70,114],"(VLMs)":[17],"across":[18,164],"various":[19],"multimodal":[20,44,95],"tasks.":[21],"However,":[22],"training":[23,49,60,181,196],"on":[24,68,182],"large-scale":[26],"datasets":[27],"is":[28,202],"computationally":[29],"expensive":[30],"and":[31,75,83,168,192],"inefficient":[32],"due":[33],"to":[34,47,104],"redundancy":[35],"in":[36,132,197],"data,":[38,191],"which":[39],"motivates":[40],"need":[42,111],"data":[45,52,96],"selection":[46,53,97,169],"improve":[48],"efficiency.":[50],"Existing":[51],"methods":[54],"VIT":[56],"either":[57],"require":[58],"costly":[59],"or":[61,71,115],"gradient":[62],"computation.":[63],"Training-free":[64],"alternatives":[65],"often":[66],"depend":[67],"proxy":[69],"datasets,":[72,167],"instruction-agnostic":[73],"representations,":[74,154],"pairwise":[76,160],"similarity":[77],"with":[78,99,102],"quadratic":[79],"complexity,":[80],"limiting":[81],"scalability":[82],"representation":[84],"fidelity.":[85],"In":[86],"this":[87],"work,":[88],"we":[89],"propose":[90],"ScalSelect,":[91],"scalable":[93,156],"training-free":[94],"method":[98],"linear-time":[100],"complexity":[101],"respect":[103],"number":[106],"samples,":[108],"eliminating":[109],"external":[113],"auxiliary":[116],"datasets.":[117],"ScalSelect":[118,173],"first":[119],"constructs":[120],"sample":[121],"representations":[122,144],"by":[123,129],"extracting":[124],"visual":[125],"features":[126],"most":[127],"attended":[128],"instruction":[130],"tokens":[131],"target":[134],"VLM,":[135],"capturing":[136],"instruction-relevant":[137],"information.":[138],"It":[139],"then":[140],"identifies":[141],"samples":[142],"whose":[143],"best":[145],"approximate":[146],"dominant":[148],"subspace":[149],"full":[152,184],"dataset":[153,185],"enabling":[155],"importance":[157],"scoring":[158],"without":[159],"comparisons.":[161],"Extensive":[162],"experiments":[163],"multiple":[165],"VLMs,":[166],"budgets":[170],"demonstrate":[171],"that":[172],"achieves":[174],"over":[175],"97.5%":[176],"using":[186],"only":[187],"16%":[188],"even":[193],"outperforms":[194],"full-data":[195],"some":[198],"settings.":[199],"The":[200],"code":[201],"available":[203],"at":[204],"\\href{https://github.com/ChangtiWu/ScalSelect}{ScalSelect}.":[205]},"counts_by_year":[],"updated_date":"2026-02-14T06:27:19.455381","created_date":"2026-02-14T00:00:00"}