{"id":"https://openalex.org/W7138348128","doi":"https://doi.org/10.48550/arxiv.2603.13346","title":"Post Training Quantization for Efficient Dataset Condensation","display_name":"Post Training Quantization for Efficient Dataset Condensation","publication_year":2026,"publication_date":"2026-03-07","ids":{"openalex":"https://openalex.org/W7138348128","doi":"https://doi.org/10.48550/arxiv.2603.13346"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2603.13346","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2603.13346","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.2603.13346","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5065935478","display_name":"Linh-Tam Tran","orcid":"https://orcid.org/0000-0002-9699-1747"},"institutions":[],"countries":[],"is_corresponding":true,"raw_author_name":"Tran, Linh-Tam","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5005200603","display_name":"Sung\u2010Ho Bae","orcid":"https://orcid.org/0000-0003-2677-3186"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Bae, Sung-Ho","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":2,"corresponding_author_ids":["https://openalex.org/A5065935478"],"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/T10036","display_name":"Advanced Neural Network Applications","score":0.5497999787330627,"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/T10036","display_name":"Advanced Neural Network Applications","score":0.5497999787330627,"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/T10775","display_name":"Generative Adversarial Networks and Image Synthesis","score":0.2558000087738037,"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":0.019200000911951065,"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/quantization","display_name":"Quantization (signal processing)","score":0.8105000257492065},{"id":"https://openalex.org/keywords/cluster-analysis","display_name":"Cluster analysis","score":0.6103000044822693},{"id":"https://openalex.org/keywords/overhead","display_name":"Overhead (engineering)","score":0.41019999980926514},{"id":"https://openalex.org/keywords/computation","display_name":"Computation","score":0.3725000023841858},{"id":"https://openalex.org/keywords/training-set","display_name":"Training set","score":0.3716000020503998},{"id":"https://openalex.org/keywords/data-compression","display_name":"Data compression","score":0.3463999927043915},{"id":"https://openalex.org/keywords/representation","display_name":"Representation (politics)","score":0.3425999879837036}],"concepts":[{"id":"https://openalex.org/C28855332","wikidata":"https://www.wikidata.org/wiki/Q198099","display_name":"Quantization (signal processing)","level":2,"score":0.8105000257492065},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6868000030517578},{"id":"https://openalex.org/C73555534","wikidata":"https://www.wikidata.org/wiki/Q622825","display_name":"Cluster analysis","level":2,"score":0.6103000044822693},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.4803999960422516},{"id":"https://openalex.org/C2779960059","wikidata":"https://www.wikidata.org/wiki/Q7113681","display_name":"Overhead (engineering)","level":2,"score":0.41019999980926514},{"id":"https://openalex.org/C45374587","wikidata":"https://www.wikidata.org/wiki/Q12525525","display_name":"Computation","level":2,"score":0.3725000023841858},{"id":"https://openalex.org/C51632099","wikidata":"https://www.wikidata.org/wiki/Q3985153","display_name":"Training set","level":2,"score":0.3716000020503998},{"id":"https://openalex.org/C78548338","wikidata":"https://www.wikidata.org/wiki/Q2493","display_name":"Data compression","level":2,"score":0.3463999927043915},{"id":"https://openalex.org/C2776359362","wikidata":"https://www.wikidata.org/wiki/Q2145286","display_name":"Representation (politics)","level":3,"score":0.3425999879837036},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.3294000029563904},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.3239000141620636},{"id":"https://openalex.org/C93372532","wikidata":"https://www.wikidata.org/wiki/Q6552455","display_name":"Linde\u2013Buzo\u2013Gray algorithm","level":3,"score":0.31040000915527344},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.28439998626708984},{"id":"https://openalex.org/C4679612","wikidata":"https://www.wikidata.org/wiki/Q866298","display_name":"Aggregate (composite)","level":2,"score":0.2768000066280365},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.27160000801086426},{"id":"https://openalex.org/C25797200","wikidata":"https://www.wikidata.org/wiki/Q828137","display_name":"Compression ratio","level":3,"score":0.2653000056743622},{"id":"https://openalex.org/C116409475","wikidata":"https://www.wikidata.org/wiki/Q1385056","display_name":"External Data Representation","level":2,"score":0.26190000772476196},{"id":"https://openalex.org/C199833920","wikidata":"https://www.wikidata.org/wiki/Q612536","display_name":"Vector quantization","level":2,"score":0.2605000138282776}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2603.13346","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2603.13346","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.2603.13346","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2603.13346","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":{"Dataset":[0],"Condensation":[1],"(DC)":[2],"distills":[3],"knowledge":[4],"from":[5],"large":[6],"datasets":[7],"into":[8],"smaller":[9],"ones,":[10],"accelerating":[11],"training":[12,64],"and":[13,133,153,188],"reducing":[14,32,54],"storage":[15,33,55,202],"requirements.":[16],"However,":[17,66],"despite":[18],"notable":[19],"progress,":[20],"prior":[21,197],"methods":[22,214],"have":[23],"largely":[24],"overlooked":[25],"the":[26,40,87,115,148,200,209],"potential":[27],"of":[28,111,117,212],"quantization":[29,46,77,107,118,159],"for":[30,121,137,158,223],"further":[31],"costs.":[34],"In":[35],"this":[36],"paper,":[37],"we":[38,67,96,125,141],"take":[39],"first":[41],"step":[42],"to":[43,79,129,146,171],"explore":[44],"post-training":[45,101],"in":[47,53,82],"dataset":[48],"condensation,":[49],"demonstrating":[50],"its":[51],"effectiveness":[52],"size":[56],"while":[57,227],"maintaining":[58],"representation":[59,83],"quality":[60],"without":[61,233],"requiring":[62],"expensive":[63],"cost.":[65],"find":[68],"that":[69,104,167,192],"at":[70,215,225],"extremely":[71],"low":[72],"bit-widths":[73],"(e.g.,":[74,219],"2-bit),":[75],"conventional":[76],"leads":[78],"substantial":[80],"degradation":[81],"quality,":[84],"negatively":[85],"impacting":[86],"networks":[88],"trained":[89],"on":[90,230],"these":[91],"data.":[92],"To":[93,113],"address":[94],"this,":[95],"propose":[97],"a":[98,143,164],"novel":[99],"\\emph{patch-based":[100],"quantization}":[102],"approach":[103],"ensures":[105],"localized":[106],"with":[108],"minimal":[109],"loss":[110],"information.":[112],"reduce":[114],"overhead":[116],"parameters,":[119],"especially":[120],"small":[122],"patch":[123],"sizes,":[124],"employ":[126],"quantization-aware":[127],"clustering":[128],"identify":[130],"similar":[131],"patches":[132],"subsequently":[134],"aggregate":[135],"them":[136],"efficient":[138],"quantization.":[139],"Furthermore,":[140],"introduce":[142],"refinement":[144],"module":[145],"align":[147],"distribution":[149],"between":[150],"original":[151],"images":[152,173,232],"their":[154],"dequantized":[155],"counterparts,":[156],"compensating":[157],"errors.":[160],"Our":[161],"method":[162,194,206],"is":[163],"plug-and-play":[165],"framework":[166],"can":[168],"be":[169],"applied":[170],"synthetic":[172],"generated":[174],"by":[175],"various":[176],"DC":[177],"methods.":[178],"Extensive":[179],"experiments":[180],"across":[181],"diverse":[182],"benchmarks":[183],"including":[184],"CIFAR-10/100,":[185],"Tiny":[186],"ImageNet,":[187],"ImageNet":[189],"subsets":[190],"demonstrate":[191],"our":[193,205],"consistently":[195],"outperforms":[196],"works":[198],"under":[199],"same":[201],"constraints.":[203],"Notably,":[204],"nearly":[207],"\\textbf{doubles":[208],"test":[210],"accuracy}":[211],"existing":[213],"extreme":[216],"compression":[217],"regimes":[218],"26.0\\%":[220],"$\\rightarrow$":[221],"54.1\\%":[222],"DM":[224],"IPC=1),":[226],"operating":[228],"directly":[229],"2-bit":[231],"additional":[234],"distillation.":[235]},"counts_by_year":[],"updated_date":"2026-04-29T09:16:38.111599","created_date":"2026-03-18T00:00:00"}