{"id":"https://openalex.org/W7163528017","doi":"https://doi.org/10.23919/date69613.2026.11539489","title":"IterQuant: Iterative Quantization Framework for Mixed-Precision LLM Compression","display_name":"IterQuant: Iterative Quantization Framework for Mixed-Precision LLM Compression","publication_year":2026,"publication_date":"2026-04-20","ids":{"openalex":"https://openalex.org/W7163528017","doi":"https://doi.org/10.23919/date69613.2026.11539489"},"language":null,"primary_location":{"id":"doi:10.23919/date69613.2026.11539489","is_oa":false,"landing_page_url":"https://doi.org/10.23919/date69613.2026.11539489","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2026 Design, Automation & Test in Europe Conference (DATE)","raw_type":"proceedings-article"},"type":"conference-paper","indexed_in":["crossref"],"open_access":{"is_oa":false,"oa_status":"closed","oa_url":null,"any_repository_has_fulltext":false},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5134159967","display_name":"Hyungyo Jeong","orcid":null},"institutions":[{"id":"https://openalex.org/I123900574","display_name":"Pohang University of Science and Technology","ror":"https://ror.org/04xysgw12","country_code":"KR","type":"education","lineage":["https://openalex.org/I123900574"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Hyungyo Jeong","raw_affiliation_strings":["POSTECH,Department of Electrical Engineering,Pohang,Republic of Korea"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"POSTECH,Department of Electrical Engineering,Pohang,Republic of Korea","institution_ids":["https://openalex.org/I123900574"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5041027404","display_name":"Junsu Kim","orcid":"https://orcid.org/0000-0001-9583-7978"},"institutions":[{"id":"https://openalex.org/I123900574","display_name":"Pohang University of Science and Technology","ror":"https://ror.org/04xysgw12","country_code":"KR","type":"education","lineage":["https://openalex.org/I123900574"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Jiwoo Kim","raw_affiliation_strings":["POSTECH,Department of Electrical Engineering,Pohang,Republic of Korea"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"POSTECH,Department of Electrical Engineering,Pohang,Republic of Korea","institution_ids":["https://openalex.org/I123900574"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5137916610","display_name":"Hyeokjun Kwon","orcid":null},"institutions":[{"id":"https://openalex.org/I142401562","display_name":"Electronics and Telecommunications Research Institute","ror":"https://ror.org/03ysstz10","country_code":"KR","type":"facility","lineage":["https://openalex.org/I142401562","https://openalex.org/I2801339556","https://openalex.org/I4210144908","https://openalex.org/I4387152098"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Hyeokjun Kwon","raw_affiliation_strings":["ETRI,Daejeon,Republic of Korea"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"ETRI,Daejeon,Republic of Korea","institution_ids":["https://openalex.org/I142401562"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100331196","display_name":"Jaeho Lee","orcid":"https://orcid.org/0000-0002-2735-0647"},"institutions":[{"id":"https://openalex.org/I123900574","display_name":"Pohang University of Science and Technology","ror":"https://ror.org/04xysgw12","country_code":"KR","type":"education","lineage":["https://openalex.org/I123900574"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Jaeho Lee","raw_affiliation_strings":["POSTECH,Department of Electrical Engineering,Pohang,Republic of Korea"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"POSTECH,Department of Electrical Engineering,Pohang,Republic of Korea","institution_ids":["https://openalex.org/I123900574"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5016264998","display_name":"Y K Lee","orcid":null},"institutions":[{"id":"https://openalex.org/I157485424","display_name":"Korea Advanced Institute of Science and Technology","ror":"https://ror.org/05apxxy63","country_code":"KR","type":"education","lineage":["https://openalex.org/I157485424"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Youngjoo Lee","raw_affiliation_strings":["KAIST,School of Electrical Engineering,Daejeon,Republic of Korea"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"KAIST,School of Electrical Engineering,Daejeon,Republic of Korea","institution_ids":["https://openalex.org/I157485424"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":3,"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":"1","last_page":"7"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T11269","display_name":"Algorithms and Data Compression","score":0.24940000474452972,"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/T11269","display_name":"Algorithms and Data Compression","score":0.24940000474452972,"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/T10901","display_name":"Advanced Data Compression Techniques","score":0.23999999463558197,"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/T12326","display_name":"Network Packet Processing and Optimization","score":0.12380000203847885,"subfield":{"id":"https://openalex.org/subfields/1708","display_name":"Hardware and Architecture"},"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/compression","display_name":"Compression (physics)","score":0.42570000886917114},{"id":"https://openalex.org/keywords/iterative-method","display_name":"Iterative method","score":0.41620001196861267},{"id":"https://openalex.org/keywords/quantization","display_name":"Quantization (signal processing)","score":0.3846000134944916},{"id":"https://openalex.org/keywords/data-compression","display_name":"Data compression","score":0.38179999589920044},{"id":"https://openalex.org/keywords/vector-quantization","display_name":"Vector quantization","score":0.30550000071525574}],"concepts":[{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.5493999719619751},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.5034000277519226},{"id":"https://openalex.org/C180016635","wikidata":"https://www.wikidata.org/wiki/Q2712821","display_name":"Compression (physics)","level":2,"score":0.42570000886917114},{"id":"https://openalex.org/C159694833","wikidata":"https://www.wikidata.org/wiki/Q2321565","display_name":"Iterative method","level":2,"score":0.41620001196861267},{"id":"https://openalex.org/C28855332","wikidata":"https://www.wikidata.org/wiki/Q198099","display_name":"Quantization (signal processing)","level":2,"score":0.3846000134944916},{"id":"https://openalex.org/C78548338","wikidata":"https://www.wikidata.org/wiki/Q2493","display_name":"Data compression","level":2,"score":0.38179999589920044},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.3517000079154968},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.31200000643730164},{"id":"https://openalex.org/C199833920","wikidata":"https://www.wikidata.org/wiki/Q612536","display_name":"Vector quantization","level":2,"score":0.30550000071525574},{"id":"https://openalex.org/C2779343474","wikidata":"https://www.wikidata.org/wiki/Q3109175","display_name":"Context (archaeology)","level":2,"score":0.2728999853134155}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.23919/date69613.2026.11539489","is_oa":false,"landing_page_url":"https://doi.org/10.23919/date69613.2026.11539489","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2026 Design, Automation & Test in Europe Conference (DATE)","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[],"awards":[],"funders":[{"id":"https://openalex.org/F4320322120","display_name":"National Research Foundation of Korea","ror":"https://ror.org/013aysd81"}],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":17,"referenced_works":["https://openalex.org/W2725159389","https://openalex.org/W2890894339","https://openalex.org/W2982041622","https://openalex.org/W2982479999","https://openalex.org/W2998617917","https://openalex.org/W3127215335","https://openalex.org/W3137147200","https://openalex.org/W3194676777","https://openalex.org/W4312310654","https://openalex.org/W4393145114","https://openalex.org/W4393147284","https://openalex.org/W4409248406","https://openalex.org/W4415796210","https://openalex.org/W4415798496","https://openalex.org/W7133198810","https://openalex.org/W7133216168","https://openalex.org/W7133245307"],"related_works":[],"abstract_inverted_index":{"Mixed-precision":[0],"quantization":[1,59,85,115,134],"is":[2],"a":[3],"promising":[4],"approach":[5],"for":[6,20],"compressing":[7],"large":[8],"language":[9],"models":[10,105],"(LLMs)":[11],"while":[12],"maintaining":[13],"output":[14,144],"quality.":[15],"However,":[16],"the":[17,64,118],"design":[18],"space":[19,66],"selecting":[21],"resolutions":[22],"of":[23],"different":[24],"layers":[25],"makes":[26],"exhaustive":[27,69],"search":[28],"intractable.":[29],"Existing":[30],"methods":[31],"either":[32],"rely":[33],"on":[34,46],"rigid":[35],"bit-width":[36],"allocation":[37,101],"schemes":[38],"or":[39],"require":[40],"extensive":[41],"hyperparameter":[42],"tuning,":[43],"often":[44],"based":[45],"inaccurate":[47],"layer-wise":[48],"sensitivity":[49,102],"metrics.":[50],"In":[51],"this":[52],"work,":[53],"we":[54],"propose":[55],"IterQuant,":[56],"an":[57],"iterative":[58],"framework":[60],"that":[61,98,128],"efficiently":[62],"explores":[63],"mixed-precision":[65,133],"without":[67],"requiring":[68],"enumeration.":[70],"By":[71],"incorporating":[72],"momentum-based":[73],"scoring":[74],"to":[75,83,108],"reflect":[76],"historical":[77],"performance":[78],"trends":[79],"and":[80,93],"parameter":[81],"grouping":[82],"balance":[84],"granularity,":[86],"IterQuant":[87,111,129],"achieves":[88],"favorable":[89],"trade-offs":[90],"between":[91],"compression":[92],"accuracy.":[94],"Unlike":[95],"prior":[96],"approaches":[97,135],"assume":[99],"bit":[100],"from":[103],"full-precision":[104],"directly":[106],"transfers":[107],"quantized":[109],"models,":[110],"dynamically":[112],"updates":[113],"its":[114],"decisions":[116],"as":[117],"model":[119],"evolves,":[120],"better":[121],"capturing":[122],"inter-layer":[123],"dependencies.":[124],"Experimental":[125],"results":[126],"demonstrate":[127],"significantly":[130],"outperforms":[131],"state-of-the-art":[132],"by":[136],"2.8%":[137],"near":[138],"4":[139],"bits":[140],"in":[141],"preserving":[142],"token-level":[143],"quality":[145],"across":[146],"various":[147],"LLM":[148],"benchmarks.":[149]},"counts_by_year":[],"updated_date":"2026-07-14T23:27:15.235271","created_date":"2026-06-05T00:00:00"}