{"id":"https://openalex.org/W7138447197","doi":"https://doi.org/10.48550/arxiv.2603.14536","title":"Distilling Latent Manifolds: Resolution Extrapolation by Variational Autoencoders","display_name":"Distilling Latent Manifolds: Resolution Extrapolation by Variational Autoencoders","publication_year":2026,"publication_date":"2026-03-15","ids":{"openalex":"https://openalex.org/W7138447197","doi":"https://doi.org/10.48550/arxiv.2603.14536"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2603.14536","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2603.14536","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.2603.14536","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5129738431","display_name":"Jiaming Chu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Chu, Jiaming","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5129694912","display_name":"Tao Wang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wang, Tao","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5129739430","display_name":"Lei Jin","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Jin, Lei","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":0.9591000080108643,"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":0.9591000080108643,"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/T11448","display_name":"Face recognition and analysis","score":0.006200000178068876,"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":0.003800000064074993,"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/autoencoder","display_name":"Autoencoder","score":0.7882999777793884},{"id":"https://openalex.org/keywords/encoder","display_name":"Encoder","score":0.6996999979019165},{"id":"https://openalex.org/keywords/extrapolation","display_name":"Extrapolation","score":0.6601999998092651},{"id":"https://openalex.org/keywords/upsampling","display_name":"Upsampling","score":0.6047999858856201},{"id":"https://openalex.org/keywords/encoding","display_name":"Encoding (memory)","score":0.536899983882904},{"id":"https://openalex.org/keywords/distillation","display_name":"Distillation","score":0.5116999745368958},{"id":"https://openalex.org/keywords/image","display_name":"Image (mathematics)","score":0.4828999936580658},{"id":"https://openalex.org/keywords/resolution","display_name":"Resolution (logic)","score":0.4471000134944916}],"concepts":[{"id":"https://openalex.org/C101738243","wikidata":"https://www.wikidata.org/wiki/Q786435","display_name":"Autoencoder","level":3,"score":0.7882999777793884},{"id":"https://openalex.org/C118505674","wikidata":"https://www.wikidata.org/wiki/Q42586063","display_name":"Encoder","level":2,"score":0.6996999979019165},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6976000070571899},{"id":"https://openalex.org/C132459708","wikidata":"https://www.wikidata.org/wiki/Q744069","display_name":"Extrapolation","level":2,"score":0.6601999998092651},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.6504999995231628},{"id":"https://openalex.org/C110384440","wikidata":"https://www.wikidata.org/wiki/Q1143270","display_name":"Upsampling","level":3,"score":0.6047999858856201},{"id":"https://openalex.org/C125411270","wikidata":"https://www.wikidata.org/wiki/Q18653","display_name":"Encoding (memory)","level":2,"score":0.536899983882904},{"id":"https://openalex.org/C204030448","wikidata":"https://www.wikidata.org/wiki/Q101017","display_name":"Distillation","level":2,"score":0.5116999745368958},{"id":"https://openalex.org/C115961682","wikidata":"https://www.wikidata.org/wiki/Q860623","display_name":"Image (mathematics)","level":2,"score":0.4828999936580658},{"id":"https://openalex.org/C138268822","wikidata":"https://www.wikidata.org/wiki/Q1051925","display_name":"Resolution (logic)","level":2,"score":0.4471000134944916},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.446399986743927},{"id":"https://openalex.org/C31972630","wikidata":"https://www.wikidata.org/wiki/Q844240","display_name":"Computer vision","level":1,"score":0.4117000102996826},{"id":"https://openalex.org/C160633673","wikidata":"https://www.wikidata.org/wiki/Q355198","display_name":"Pixel","level":2,"score":0.40450000762939453},{"id":"https://openalex.org/C141379421","wikidata":"https://www.wikidata.org/wiki/Q6094427","display_name":"Iterative reconstruction","level":2,"score":0.3889999985694885},{"id":"https://openalex.org/C11727466","wikidata":"https://www.wikidata.org/wiki/Q1628157","display_name":"Inpainting","level":3,"score":0.376800000667572},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.37470000982284546},{"id":"https://openalex.org/C2780586882","wikidata":"https://www.wikidata.org/wiki/Q7520643","display_name":"Simple (philosophy)","level":2,"score":0.359499990940094},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.34769999980926514},{"id":"https://openalex.org/C39890363","wikidata":"https://www.wikidata.org/wiki/Q36108","display_name":"Generative grammar","level":2,"score":0.31189998984336853},{"id":"https://openalex.org/C167966045","wikidata":"https://www.wikidata.org/wiki/Q5532625","display_name":"Generative model","level":3,"score":0.29670000076293945},{"id":"https://openalex.org/C108583219","wikidata":"https://www.wikidata.org/wiki/Q197536","display_name":"Deep learning","level":2,"score":0.29440000653266907},{"id":"https://openalex.org/C205372480","wikidata":"https://www.wikidata.org/wiki/Q210521","display_name":"Image resolution","level":2,"score":0.27639999985694885},{"id":"https://openalex.org/C3020199158","wikidata":"https://www.wikidata.org/wiki/Q210521","display_name":"High resolution","level":2,"score":0.25529998540878296}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2603.14536","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2603.14536","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.2603.14536","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2603.14536","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":{"Variational":[0],"Autoencoder":[1],"(VAE)":[2],"encoders":[3],"play":[4],"a":[5,56,63,201],"critical":[6],"role":[7],"in":[8,59,225],"modern":[9],"generative":[10],"models,":[11],"yet":[12],"their":[13,43],"computational":[14],"cost":[15,188],"often":[16],"motivates":[17],"the":[18,34,39,98,109,132,185,212,218,222],"use":[19],"of":[20,221],"knowledge":[21,220],"distillation":[22,171,213],"or":[23],"quantification":[24],"to":[25,42,103,154],"obtain":[26],"compact":[27,64],"alternatives.":[28],"Existing":[29],"studies":[30],"typically":[31],"believe":[32],"that":[33,122,142,168,184],"model":[35,214,224],"work":[36],"better":[37],"on":[38,138,189],"samples":[40],"closed":[41],"training":[44,187],"data":[45,49],"distribution":[46],"than":[47,177],"unseen":[48,88],"distribution.":[50],"In":[51],"this":[52],"work,":[53],"we":[54,140],"report":[55],"counter-intuitive":[57],"phenomenon":[58],"VAE":[60,169,202],"encoder":[61,65,100,170],"distillation:":[62],"distilled":[66,99],"only":[67],"at":[68,75,86],"low":[69,209],"resolutions":[70,119],"exhibits":[71],"poor":[72],"reconstruction":[73,206],"performance":[74],"its":[76],"native":[77],"resolution,":[78,97],"but":[79],"achieves":[80],"dramatically":[81],"improved":[82],"results":[83],"when":[84],"evaluated":[85],"higher,":[87],"input":[89],"resolutions.":[90],"Despite":[91],"never":[92],"being":[93],"trained":[94],"beyond":[95],"$256^2$":[96],"generalizes":[101],"effectively":[102],"$512^2$":[104],"resolution":[105,112,144,210],"inputs,":[106],"partially":[107],"inheriting":[108],"teacher":[110,223],"model's":[111],"preference.We":[113],"further":[114],"analyze":[115],"latent":[116,126,174],"distributions":[117],"across":[118,157],"and":[120,149,162,192],"find":[121],"higher-resolution":[123],"inputs":[124,146],"produce":[125],"representations":[127],"more":[128],"closely":[129],"aligned":[130],"with":[131,203],"teacher's":[133],"manifold.":[134],"Through":[135],"extensive":[136],"experiments":[137],"ImageNet-256,":[139],"show":[141],"simple":[143],"remapping-upsampling":[145],"before":[147],"encoding":[148],"downsampling":[150],"reconstructions":[151],"for":[152,199],"evaluation-leads":[153],"substantial":[155],"gains":[156],"PSNR,":[158],"MSE,":[159],"SSIM,":[160],"LPIPS,":[161],"rFID":[163],"metrics.":[164],"These":[165],"findings":[166],"suggest":[167],"learns":[172],"resolution-consistent":[173],"manifolds":[175],"rather":[176],"resolution-specific":[178],"pixel":[179],"mappings.":[180],"This":[181],"also":[182],"means":[183],"high":[186],"memory,":[190],"time":[191],"high-resolution":[193,204,226],"datasets":[194],"are":[195],"not":[196],"necessary":[197],"conditions":[198],"distilling":[200],"image":[205,227],"capabilities.":[207],"On":[208],"datasets,":[211],"still":[215],"could":[216],"learn":[217],"detailed":[219],"reconstruction.":[228]},"counts_by_year":[],"updated_date":"2026-07-28T07:46:37.118299","created_date":"2026-03-18T00:00:00"}