{"id":"https://openalex.org/W4411727432","doi":"https://doi.org/10.1109/iscas56072.2025.11043993","title":"An Integer-only Quantization Framework for Edge Deployment of Large Language Models","display_name":"An Integer-only Quantization Framework for Edge Deployment of Large Language Models","publication_year":2025,"publication_date":"2025-05-25","ids":{"openalex":"https://openalex.org/W4411727432","doi":"https://doi.org/10.1109/iscas56072.2025.11043993"},"language":"en","primary_location":{"id":"doi:10.1109/iscas56072.2025.11043993","is_oa":false,"landing_page_url":"https://doi.org/10.1109/iscas56072.2025.11043993","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2025 IEEE International Symposium on Circuits and Systems (ISCAS)","raw_type":"proceedings-article"},"type":"article","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/A5104304912","display_name":"Yaqi Hu","orcid":"https://orcid.org/0000-0003-2597-1332"},"institutions":[{"id":"https://openalex.org/I99065089","display_name":"Tsinghua University","ror":"https://ror.org/03cve4549","country_code":"CN","type":"education","lineage":["https://openalex.org/I99065089"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Yaqi Hu","raw_affiliation_strings":["Tsinghua University,Shenzhen International Graduate School,Shenzhen,China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Tsinghua University,Shenzhen International Graduate School,Shenzhen,China","institution_ids":["https://openalex.org/I99065089"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5010267099","display_name":"Zhuqing Yuan","orcid":"https://orcid.org/0000-0002-1042-9711"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhuqing Yuan","raw_affiliation_strings":["Pi2star Technology Ltd.,Beijing,China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Pi2star Technology Ltd.,Beijing,China","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5049018667","display_name":"Weichen Gao","orcid":null},"institutions":[{"id":"https://openalex.org/I99065089","display_name":"Tsinghua University","ror":"https://ror.org/03cve4549","country_code":"CN","type":"education","lineage":["https://openalex.org/I99065089"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Weichen Gao","raw_affiliation_strings":["Tsinghua University,Department of Electronic Engineering,Beijing,China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Tsinghua University,Department of Electronic Engineering,Beijing,China","institution_ids":["https://openalex.org/I99065089"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100600225","display_name":"Sheng Zhang","orcid":"https://orcid.org/0000-0001-8059-2269"},"institutions":[{"id":"https://openalex.org/I99065089","display_name":"Tsinghua University","ror":"https://ror.org/03cve4549","country_code":"CN","type":"education","lineage":["https://openalex.org/I99065089"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Sheng Zhang","raw_affiliation_strings":["Tsinghua University,Shenzhen International Graduate School,Shenzhen,China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Tsinghua University,Shenzhen International Graduate School,Shenzhen,China","institution_ids":["https://openalex.org/I99065089"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5045721867","display_name":"Yongpan Liu","orcid":"https://orcid.org/0000-0002-4892-2309"},"institutions":[{"id":"https://openalex.org/I99065089","display_name":"Tsinghua University","ror":"https://ror.org/03cve4549","country_code":"CN","type":"education","lineage":["https://openalex.org/I99065089"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Yongpan Liu","raw_affiliation_strings":["Tsinghua University,Department of Electronic Engineering,Beijing,China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Tsinghua University,Department of Electronic Engineering,Beijing,China","institution_ids":["https://openalex.org/I99065089"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":5,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":2.2813,"has_fulltext":false,"cited_by_count":1,"citation_normalized_percentile":{"value":0.88981601,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":95,"max":98},"biblio":{"volume":null,"issue":null,"first_page":"1","last_page":"5"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10203","display_name":"Recommender Systems and Techniques","score":0.739799976348877,"subfield":{"id":"https://openalex.org/subfields/1710","display_name":"Information Systems"},"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/T10203","display_name":"Recommender Systems and Techniques","score":0.739799976348877,"subfield":{"id":"https://openalex.org/subfields/1710","display_name":"Information Systems"},"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/T10715","display_name":"Distributed and Parallel Computing Systems","score":0.6865000128746033,"subfield":{"id":"https://openalex.org/subfields/1705","display_name":"Computer Networks and Communications"},"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/T11273","display_name":"Advanced Graph Neural Networks","score":0.6536999940872192,"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/computer-science","display_name":"Computer science","score":0.699161946773529},{"id":"https://openalex.org/keywords/software-deployment","display_name":"Software deployment","score":0.6826623678207397},{"id":"https://openalex.org/keywords/quantization","display_name":"Quantization (signal processing)","score":0.6001869440078735},{"id":"https://openalex.org/keywords/integer","display_name":"Integer (computer science)","score":0.5269442796707153},{"id":"https://openalex.org/keywords/enhanced-data-rates-for-gsm-evolution","display_name":"Enhanced Data Rates for GSM Evolution","score":0.4763748347759247},{"id":"https://openalex.org/keywords/theoretical-computer-science","display_name":"Theoretical computer science","score":0.3457910418510437},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.2586248517036438},{"id":"https://openalex.org/keywords/algorithm","display_name":"Algorithm","score":0.24250108003616333},{"id":"https://openalex.org/keywords/programming-language","display_name":"Programming language","score":0.24042734503746033},{"id":"https://openalex.org/keywords/software-engineering","display_name":"Software engineering","score":0.12617531418800354}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.699161946773529},{"id":"https://openalex.org/C105339364","wikidata":"https://www.wikidata.org/wiki/Q2297740","display_name":"Software deployment","level":2,"score":0.6826623678207397},{"id":"https://openalex.org/C28855332","wikidata":"https://www.wikidata.org/wiki/Q198099","display_name":"Quantization (signal processing)","level":2,"score":0.6001869440078735},{"id":"https://openalex.org/C97137487","wikidata":"https://www.wikidata.org/wiki/Q729138","display_name":"Integer (computer science)","level":2,"score":0.5269442796707153},{"id":"https://openalex.org/C162307627","wikidata":"https://www.wikidata.org/wiki/Q204833","display_name":"Enhanced Data Rates for GSM Evolution","level":2,"score":0.4763748347759247},{"id":"https://openalex.org/C80444323","wikidata":"https://www.wikidata.org/wiki/Q2878974","display_name":"Theoretical computer science","level":1,"score":0.3457910418510437},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.2586248517036438},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.24250108003616333},{"id":"https://openalex.org/C199360897","wikidata":"https://www.wikidata.org/wiki/Q9143","display_name":"Programming language","level":1,"score":0.24042734503746033},{"id":"https://openalex.org/C115903868","wikidata":"https://www.wikidata.org/wiki/Q80993","display_name":"Software engineering","level":1,"score":0.12617531418800354}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/iscas56072.2025.11043993","is_oa":false,"landing_page_url":"https://doi.org/10.1109/iscas56072.2025.11043993","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2025 IEEE International Symposium on Circuits and Systems (ISCAS)","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[],"awards":[],"funders":[{"id":"https://openalex.org/F4320326444","display_name":"Nova","ror":null}],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":21,"referenced_works":["https://openalex.org/W2093647425","https://openalex.org/W3196923642","https://openalex.org/W3211525823","https://openalex.org/W3213539149","https://openalex.org/W4286571858","https://openalex.org/W4293025165","https://openalex.org/W4308083526","https://openalex.org/W4389524393","https://openalex.org/W6727099177","https://openalex.org/W6769627184","https://openalex.org/W6788001715","https://openalex.org/W6838633097","https://openalex.org/W6842758154","https://openalex.org/W6847478871","https://openalex.org/W6851828392","https://openalex.org/W6856414091","https://openalex.org/W6856696905","https://openalex.org/W6856969750","https://openalex.org/W6857799723","https://openalex.org/W6857951981","https://openalex.org/W6877022392"],"related_works":["https://openalex.org/W2770234245","https://openalex.org/W96612179","https://openalex.org/W4229499248","https://openalex.org/W2566006169","https://openalex.org/W1567818861","https://openalex.org/W2987774938","https://openalex.org/W4256492088","https://openalex.org/W632915154","https://openalex.org/W2055733372","https://openalex.org/W3022067003"],"abstract_inverted_index":{"The":[0],"rapid":[1],"growth":[2],"in":[3],"the":[4,74,113,125],"parameter":[5],"size":[6,102],"of":[7],"large":[8,65],"language":[9,110],"models":[10],"(LLMs)":[11],"has":[12],"introduced":[13],"significant":[14],"challenges":[15],"for":[16,36,76,87,96],"deployment":[17,38],"on":[18,28,39,51,108],"edge":[19,40,88],"devices.":[20,41],"To":[21],"address":[22],"these":[23],"challenges,":[24],"this":[25],"paper":[26],"focuses":[27],"developing":[29],"a":[30],"post-training":[31],"quantization":[32,61,77,95],"(PTQ)":[33],"framework":[34,43,120],"tailored":[35],"LLM":[37],"Our":[42,90],"introduces":[44],"an":[45],"enhanced":[46],"channel":[47,52,57,66],"smoothing":[48],"technique":[49],"based":[50],"value":[53,67],"ranges,":[54],"combined":[55],"with":[56,64],"reordering":[58],"to":[59],"mitigate":[60],"errors":[62],"associated":[63],"range":[68],"differences":[69],"and":[70,78],"activation":[71],"outliers,":[72],"reducing":[73,100],"need":[75],"dequantization":[79],"steps":[80],"during":[81],"inference,":[82],"making":[83],"it":[84],"more":[85],"suitable":[86],"deployment.":[89],"approach":[91],"achieves":[92],"full":[93],"integer":[94],"all":[97],"model":[98,101],"operations,":[99],"by":[103],"4\u00d7.":[104],"Through":[105],"extensive":[106],"experiments":[107],"various":[109],"tasks":[111],"using":[112],"OPT":[114],"model,":[115],"we":[116],"demonstrate":[117],"that":[118],"our":[119],"surpasses":[121],"state-of-the-art":[122],"methods":[123],"under":[124],"W4A4":[126],"configuration.":[127]},"counts_by_year":[{"year":2026,"cited_by_count":1}],"updated_date":"2026-06-19T17:40:00.097472","created_date":"2025-10-10T00:00:00"}