{"id":"https://openalex.org/W4410582506","doi":"https://doi.org/10.23919/date64628.2025.10992912","title":"Cocktail: Chunk-Adaptive Mixed-Precision Quantization for Long-Context LLM Inference","display_name":"Cocktail: Chunk-Adaptive Mixed-Precision Quantization for Long-Context LLM Inference","publication_year":2025,"publication_date":"2025-03-31","ids":{"openalex":"https://openalex.org/W4410582506","doi":"https://doi.org/10.23919/date64628.2025.10992912"},"language":"en","primary_location":{"id":"doi:10.23919/date64628.2025.10992912","is_oa":false,"landing_page_url":"https://doi.org/10.23919/date64628.2025.10992912","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2025 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/A5101714910","display_name":"Wei Tao","orcid":"https://orcid.org/0009-0004-2928-3451"},"institutions":[{"id":"https://openalex.org/I47720641","display_name":"Huazhong University of Science and Technology","ror":"https://ror.org/00p991c53","country_code":"CN","type":"education","lineage":["https://openalex.org/I47720641"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Wei Tao","raw_affiliation_strings":["Huazhong University of Science and Technology,Wuhan,China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Huazhong University of Science and Technology,Wuhan,China","institution_ids":["https://openalex.org/I47720641"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100392907","display_name":"Bin Zhang","orcid":"https://orcid.org/0000-0003-1323-8678"},"institutions":[{"id":"https://openalex.org/I47720641","display_name":"Huazhong University of Science and Technology","ror":"https://ror.org/00p991c53","country_code":"CN","type":"education","lineage":["https://openalex.org/I47720641"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Bin Zhang","raw_affiliation_strings":["Huazhong University of Science and Technology,Wuhan,China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Huazhong University of Science and Technology,Wuhan,China","institution_ids":["https://openalex.org/I47720641"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100940847","display_name":"Xiaoyang Qu","orcid":"https://orcid.org/0009-0009-6311-4332"},"institutions":[{"id":"https://openalex.org/I4210152380","display_name":"Shenzhen Technology University","ror":"https://ror.org/04qzpec27","country_code":"CN","type":"education","lineage":["https://openalex.org/I4210152380"]},{"id":"https://openalex.org/I4401726822","display_name":"Ping An (China)","ror":"https://ror.org/004yv2z91","country_code":null,"type":"company","lineage":["https://openalex.org/I4401726822"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Xiaoyang Qu","raw_affiliation_strings":["Ping An Technology (Shenzhen) Co., Ltd,Shenzhen,China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Ping An Technology (Shenzhen) Co., Ltd,Shenzhen,China","institution_ids":["https://openalex.org/I4210152380","https://openalex.org/I4401726822"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5044866795","display_name":"Jiguang Wan","orcid":"https://orcid.org/0000-0003-3440-4460"},"institutions":[{"id":"https://openalex.org/I47720641","display_name":"Huazhong University of Science and Technology","ror":"https://ror.org/00p991c53","country_code":"CN","type":"education","lineage":["https://openalex.org/I47720641"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Jiguang Wan","raw_affiliation_strings":["Huazhong University of Science and Technology,Wuhan,China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Huazhong University of Science and Technology,Wuhan,China","institution_ids":["https://openalex.org/I47720641"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5074472751","display_name":"Jianzong Wang","orcid":"https://orcid.org/0000-0002-9237-4231"},"institutions":[{"id":"https://openalex.org/I4210152380","display_name":"Shenzhen Technology University","ror":"https://ror.org/04qzpec27","country_code":"CN","type":"education","lineage":["https://openalex.org/I4210152380"]},{"id":"https://openalex.org/I4401726822","display_name":"Ping An (China)","ror":"https://ror.org/004yv2z91","country_code":null,"type":"company","lineage":["https://openalex.org/I4401726822"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Jianzong Wang","raw_affiliation_strings":["Ping An Technology (Shenzhen) Co., Ltd,Shenzhen,China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Ping An Technology (Shenzhen) Co., Ltd,Shenzhen,China","institution_ids":["https://openalex.org/I4210152380","https://openalex.org/I4401726822"]}]}],"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":3,"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/T10181","display_name":"Natural Language Processing Techniques","score":0.9769999980926514,"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/T10181","display_name":"Natural Language Processing Techniques","score":0.9769999980926514,"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":0.7332172393798828},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.6978359818458557},{"id":"https://openalex.org/keywords/quantization","display_name":"Quantization (signal processing)","score":0.6259633898735046},{"id":"https://openalex.org/keywords/context","display_name":"Context (archaeology)","score":0.5140146017074585},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.47023317217826843},{"id":"https://openalex.org/keywords/machine-learning","display_name":"Machine learning","score":0.34482860565185547},{"id":"https://openalex.org/keywords/algorithm","display_name":"Algorithm","score":0.27443835139274597}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7332172393798828},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.6978359818458557},{"id":"https://openalex.org/C28855332","wikidata":"https://www.wikidata.org/wiki/Q198099","display_name":"Quantization (signal processing)","level":2,"score":0.6259633898735046},{"id":"https://openalex.org/C2779343474","wikidata":"https://www.wikidata.org/wiki/Q3109175","display_name":"Context (archaeology)","level":2,"score":0.5140146017074585},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.47023317217826843},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.34482860565185547},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.27443835139274597},{"id":"https://openalex.org/C151730666","wikidata":"https://www.wikidata.org/wiki/Q7205","display_name":"Paleontology","level":1,"score":0.0},{"id":"https://openalex.org/C86803240","wikidata":"https://www.wikidata.org/wiki/Q420","display_name":"Biology","level":0,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.23919/date64628.2025.10992912","is_oa":false,"landing_page_url":"https://doi.org/10.23919/date64628.2025.10992912","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2025 Design, Automation & Test in Europe Conference (DATE)","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":14,"referenced_works":["https://openalex.org/W3180230246","https://openalex.org/W4387321091","https://openalex.org/W4389518760","https://openalex.org/W4401568701","https://openalex.org/W4402671766","https://openalex.org/W6739901393","https://openalex.org/W6777615688","https://openalex.org/W6838322825","https://openalex.org/W6847478871","https://openalex.org/W6850927664","https://openalex.org/W6850928896","https://openalex.org/W6853192989","https://openalex.org/W6857799723","https://openalex.org/W6877022392"],"related_works":["https://openalex.org/W2961085424","https://openalex.org/W4306674287","https://openalex.org/W4387369504","https://openalex.org/W4394896187","https://openalex.org/W3170094116","https://openalex.org/W4386462264","https://openalex.org/W3107602296","https://openalex.org/W4364306694","https://openalex.org/W4312192474","https://openalex.org/W4283697347"],"abstract_inverted_index":{"Recently,":[0],"large":[1],"language":[2],"models":[3,156],"(LLMs)":[4],"have":[5],"been":[6],"able":[7],"to":[8,35,71],"handle":[9],"longer":[10,12],"and":[11,26,53,84,113,157],"contexts.":[13],"However,":[14],"a":[15,61],"context":[16,111],"that":[17,146],"is":[18,47,161],"too":[19],"long":[20],"may":[21],"cause":[22],"intolerant":[23],"inference":[24,141],"latency":[25],"GPU":[27],"memory":[28],"usage.":[29],"Existing":[30],"methods":[31,153],"propose":[32],"mixed-precision":[33,69,138],"quantization":[34,70,82,90,139,152],"the":[36,50,73,93,98,105,109,114,117,126,133],"key-value":[37],"(KV)":[38],"cache":[39,87,100,123,128,151],"in":[40,49,140],"LLMs":[41],"based":[42,103],"on":[43,104,154,163],"token":[44],"granularity,":[45],"which":[46,66],"time-consuming":[48],"search":[51,83,91],"process":[52],"hardware":[54,134],"inefficient":[55],"during":[56],"computation.":[57,88,142],"This":[58],"paper":[59],"introduces":[60],"novel":[62],"approach":[63],"called":[64],"Cocktail,":[65],"employs":[67],"chunk-adaptive":[68],"optimize":[72],"KV":[74,86,99,122,127,150],"cache.":[75],"Cocktail":[76,147],"consists":[77],"of":[78,97],"two":[79],"modules:":[80],"chunk-level":[81,85,121],"Chunk-level":[89],"determines":[92],"optimal":[94],"bitwidth":[95],"configuration":[96],"chunks":[101,112,129],"quickly":[102],"similarity":[106],"scores":[107],"between":[108],"corresponding":[110],"query,":[115],"maintaining":[116],"model":[118],"accuracy.":[119],"Furthermore,":[120],"computation":[124],"reorders":[125],"before":[130],"quantization,":[131],"avoiding":[132],"inefficiency":[135],"caused":[136],"by":[137],"Extensive":[143],"experiments":[144],"demonstrate":[145],"outperforms":[148],"state-of-the-art":[149],"various":[155],"datasets.":[158],"Our":[159],"code":[160],"presented":[162],"https://github.com/Sullivan12138/Cocktail.":[164]},"counts_by_year":[{"year":2026,"cited_by_count":2},{"year":2025,"cited_by_count":1}],"updated_date":"2026-07-14T23:27:15.235271","created_date":"2025-10-10T00:00:00"}