{"id":"https://openalex.org/W7161764412","doi":"https://doi.org/10.48550/arxiv.2605.18836","title":"Spectral Gradient Surgery for Domain-Generalizable Dataset Distillation","display_name":"Spectral Gradient Surgery for Domain-Generalizable Dataset Distillation","publication_year":2026,"publication_date":"2026-05-13","ids":{"openalex":"https://openalex.org/W7161764412","doi":"https://doi.org/10.48550/arxiv.2605.18836"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2605.18836","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.18836","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":"Preprint"},"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.48550/arxiv.2605.18836","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5136519281","display_name":"Minyoung Oh","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Oh, Minyoung","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5115904629","display_name":"Najeong Chae","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Chae, Najeong","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5136543628","display_name":"Jae-Young Sim","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Sim, Jae-Young","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/T11307","display_name":"Domain Adaptation and Few-Shot Learning","score":0.9438999891281128,"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/T11307","display_name":"Domain Adaptation and Few-Shot Learning","score":0.9438999891281128,"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/T11689","display_name":"Adversarial Robustness in Machine Learning","score":0.013199999928474426,"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/T11652","display_name":"Imbalanced Data Classification Techniques","score":0.00860000029206276,"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/distillation","display_name":"Distillation","score":0.83160001039505},{"id":"https://openalex.org/keywords/generalization","display_name":"Generalization","score":0.7307000160217285},{"id":"https://openalex.org/keywords/overhead","display_name":"Overhead (engineering)","score":0.5950000286102295},{"id":"https://openalex.org/keywords/matching","display_name":"Matching (statistics)","score":0.527899980545044},{"id":"https://openalex.org/keywords/domain","display_name":"Domain (mathematical analysis)","score":0.4677000045776367},{"id":"https://openalex.org/keywords/synthetic-data","display_name":"Synthetic data","score":0.4505000114440918},{"id":"https://openalex.org/keywords/fractionating-column","display_name":"Fractionating column","score":0.3917999863624573}],"concepts":[{"id":"https://openalex.org/C204030448","wikidata":"https://www.wikidata.org/wiki/Q101017","display_name":"Distillation","level":2,"score":0.83160001039505},{"id":"https://openalex.org/C177148314","wikidata":"https://www.wikidata.org/wiki/Q170084","display_name":"Generalization","level":2,"score":0.7307000160217285},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6283000111579895},{"id":"https://openalex.org/C2779960059","wikidata":"https://www.wikidata.org/wiki/Q7113681","display_name":"Overhead (engineering)","level":2,"score":0.5950000286102295},{"id":"https://openalex.org/C165064840","wikidata":"https://www.wikidata.org/wiki/Q1321061","display_name":"Matching (statistics)","level":2,"score":0.527899980545044},{"id":"https://openalex.org/C36503486","wikidata":"https://www.wikidata.org/wiki/Q11235244","display_name":"Domain (mathematical analysis)","level":2,"score":0.4677000045776367},{"id":"https://openalex.org/C160920958","wikidata":"https://www.wikidata.org/wiki/Q7662746","display_name":"Synthetic data","level":2,"score":0.4505000114440918},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.40389999747276306},{"id":"https://openalex.org/C154030694","wikidata":"https://www.wikidata.org/wiki/Q1436074","display_name":"Fractionating column","level":3,"score":0.3917999863624573},{"id":"https://openalex.org/C62354387","wikidata":"https://www.wikidata.org/wiki/Q875399","display_name":"Boundary (topology)","level":2,"score":0.3813999891281128},{"id":"https://openalex.org/C95713431","wikidata":"https://www.wikidata.org/wiki/Q631425","display_name":"Vulnerability (computing)","level":2,"score":0.3813999891281128},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.367000013589859},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.36579999327659607},{"id":"https://openalex.org/C12725497","wikidata":"https://www.wikidata.org/wiki/Q810247","display_name":"Baseline (sea)","level":2,"score":0.35569998621940613},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.3456999957561493},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.3395000100135803},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.33550000190734863},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.33250001072883606},{"id":"https://openalex.org/C110121322","wikidata":"https://www.wikidata.org/wiki/Q865811","display_name":"Distribution (mathematics)","level":2,"score":0.31459999084472656},{"id":"https://openalex.org/C51632099","wikidata":"https://www.wikidata.org/wiki/Q3985153","display_name":"Training set","level":2,"score":0.27889999747276306},{"id":"https://openalex.org/C2777894999","wikidata":"https://www.wikidata.org/wiki/Q4781758","display_name":"Approx","level":2,"score":0.25940001010894775}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2605.18836","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.18836","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":"Preprint"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2605.18836","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.18836","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":"Preprint"},"sustainable_development_goals":[{"display_name":"Reduced inequalities","score":0.7646247148513794,"id":"https://metadata.un.org/sdg/10"}],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Dataset":[0,92],"Distillation":[1,93],"(DD)":[2],"synthesizes":[3],"a":[4,14,95,113],"compact":[5,64],"synthetic":[6,65,140],"dataset":[7,82],"that":[8,22,34,75,99,158,198,205,219],"preserves":[9],"the":[10,26,57,78,124,130,138,150,165,189,210],"training":[11,30],"utility":[12],"of":[13,60,81,105,118,127,132,155],"full":[15],"dataset.":[16,212],"However,":[17],"its":[18],"standard":[19,190],"formulation":[20],"assumes":[21],"test":[23],"data":[24],"follow":[25],"same":[27],"distribution":[28],"as":[29],"data,":[31],"an":[32],"assumption":[33],"rarely":[35],"holds":[36],"in":[37,164],"practice.":[38],"A":[39],"straightforward":[40],"extension-applying":[41],"post-hoc":[42],"Domain":[43,90],"Generalization":[44],"(DG)":[45],"techniques":[46],"to":[47,129,148],"distilled":[48,106,211],"data-is":[49],"ill-suited":[50],"because":[51],"existing":[52,230],"DG":[53],"methods":[54],"rely":[55],"on":[56,184,215],"natural":[58],"diversity":[59,208],"real":[61],"datasets,":[62],"which":[63,169,180],"sets":[66],"inherently":[67],"lack,":[68],"while":[69,225],"also":[70],"incurring":[71],"substantial":[72],"augmentation":[73],"overhead":[74],"conflicts":[76],"with":[77,193,229],"efficiency":[79],"objective":[80],"distillation.":[83],"To":[84],"address":[85],"this":[86,110,185],"limitation,":[87],"we":[88],"introduce":[89],"Generalizable":[91],"(DGDD),":[94],"new":[96],"problem":[97,111],"setting":[98],"explicitly":[100,206],"targets":[101],"out-of-distribution":[102],"(OOD)":[103],"generalization":[104,224],"datasets.":[107],"We":[108,122],"study":[109],"through":[112],"widely":[114],"adopted":[115],"DD":[116],"baseline":[117],"Distribution":[119],"Matching":[120],"(DM).":[121],"attribute":[123],"OOD":[125,223],"vulnerability":[126],"DM":[128,191,231],"entanglement":[131],"class-discriminative":[133],"and":[134,142,203],"domain-specific":[135],"information":[136],"within":[137,209],"compressed":[139],"set,":[141],"propose":[143],"Spectral":[144],"Gradient":[145],"Surgery":[146],"(SGS)":[147],"disentangle":[149],"two.":[151],"The":[152],"key":[153],"insight":[154],"SGS":[156,187,220],"is":[157],"cross-domain":[159,200],"agreement":[160],"among":[161],"domain-wise":[162],"gradients":[163],"spectral":[166],"domain":[167],"reveals":[168],"gradient":[170],"components":[171,202],"are":[172,177,181],"shared":[173,201],"across":[174],"source":[175],"domains-and":[176],"therefore":[178],"class-discriminative-and":[179],"domain-specific.":[182],"Based":[183],"observation,":[186],"augments":[188],"update":[192],"two":[194],"complementary":[195],"gradients:":[196],"one":[197],"reinforces":[199],"another":[204],"promotes":[207],"Extensive":[213],"experiments":[214],"diverse-scale":[216],"benchmarks":[217],"demonstrate":[218],"substantially":[221],"improves":[222],"remaining":[226],"plug-and-play":[227],"compatible":[228],"methods.":[232]},"counts_by_year":[],"updated_date":"2026-07-01T06:00:48.157686","created_date":"2026-05-21T00:00:00"}