{"id":"https://openalex.org/W4392944581","doi":"https://doi.org/10.1109/iceic61013.2024.10457111","title":"Searching Optimal Floating-Point Format for Sub-8-Bit Large Language Model Inference","display_name":"Searching Optimal Floating-Point Format for Sub-8-Bit Large Language Model Inference","publication_year":2024,"publication_date":"2024-01-28","ids":{"openalex":"https://openalex.org/W4392944581","doi":"https://doi.org/10.1109/iceic61013.2024.10457111"},"language":"en","primary_location":{"id":"doi:10.1109/iceic61013.2024.10457111","is_oa":false,"landing_page_url":"http://dx.doi.org/10.1109/iceic61013.2024.10457111","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2024 International Conference on Electronics, Information, and Communication (ICEIC)","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/A5051424291","display_name":"Youngdeok Hwang","orcid":"https://orcid.org/0000-0001-7563-5156"},"institutions":[{"id":"https://openalex.org/I174216632","display_name":"City University of New York","ror":"https://ror.org/00453a208","country_code":"US","type":"education","lineage":["https://openalex.org/I174216632"]}],"countries":["US"],"is_corresponding":true,"raw_author_name":"Youngdeok Hwang","raw_affiliation_strings":["City University of New York"],"affiliations":[{"raw_affiliation_string":"City University of New York","institution_ids":["https://openalex.org/I174216632"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5036363860","display_name":"Janghwan Lee","orcid":null},"institutions":[{"id":"https://openalex.org/I4575257","display_name":"Hanyang University","ror":"https://ror.org/046865y68","country_code":"KR","type":"education","lineage":["https://openalex.org/I4575257"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Janghwan Lee","raw_affiliation_strings":["Hanyang University"],"affiliations":[{"raw_affiliation_string":"Hanyang University","institution_ids":["https://openalex.org/I4575257"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101918926","display_name":"Jiwoong Park","orcid":"https://orcid.org/0000-0002-0511-1442"},"institutions":[{"id":"https://openalex.org/I4575257","display_name":"Hanyang University","ror":"https://ror.org/046865y68","country_code":"KR","type":"education","lineage":["https://openalex.org/I4575257"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Jiwoong Park","raw_affiliation_strings":["Hanyang University"],"affiliations":[{"raw_affiliation_string":"Hanyang University","institution_ids":["https://openalex.org/I4575257"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5068625256","display_name":"Jieun Lim","orcid":"https://orcid.org/0000-0002-8985-1957"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Jieun Lim","raw_affiliation_strings":["SAPEON Korea Inc"],"affiliations":[{"raw_affiliation_string":"SAPEON Korea Inc","institution_ids":[]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5101935712","display_name":"Jungwook Choi","orcid":"https://orcid.org/0000-0002-5691-4771"},"institutions":[{"id":"https://openalex.org/I4575257","display_name":"Hanyang University","ror":"https://ror.org/046865y68","country_code":"KR","type":"education","lineage":["https://openalex.org/I4575257"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Jungwook Choi","raw_affiliation_strings":["Hanyang University"],"affiliations":[{"raw_affiliation_string":"Hanyang University","institution_ids":["https://openalex.org/I4575257"]}]}],"institutions":[],"countries_distinct_count":2,"institutions_distinct_count":5,"corresponding_author_ids":["https://openalex.org/A5051424291"],"corresponding_institution_ids":["https://openalex.org/I174216632"],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":{"value":0.02709862,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"1","last_page":"4"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10201","display_name":"Speech Recognition and Synthesis","score":0.9415000081062317,"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/T10201","display_name":"Speech Recognition and Synthesis","score":0.9415000081062317,"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":0.7402946352958679},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.6544725894927979},{"id":"https://openalex.org/keywords/bit","display_name":"Bit (key)","score":0.6019614934921265},{"id":"https://openalex.org/keywords/point","display_name":"Point (geometry)","score":0.5693848133087158},{"id":"https://openalex.org/keywords/language-model","display_name":"Language model","score":0.5490116477012634},{"id":"https://openalex.org/keywords/programming-language","display_name":"Programming language","score":0.524160623550415},{"id":"https://openalex.org/keywords/floating-point","display_name":"Floating point","score":0.4209928810596466},{"id":"https://openalex.org/keywords/algorithm","display_name":"Algorithm","score":0.3864714801311493},{"id":"https://openalex.org/keywords/parallel-computing","display_name":"Parallel computing","score":0.34753692150115967},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.29931098222732544},{"id":"https://openalex.org/keywords/mathematics","display_name":"Mathematics","score":0.14535021781921387},{"id":"https://openalex.org/keywords/computer-network","display_name":"Computer network","score":0.08781862258911133}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7402946352958679},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.6544725894927979},{"id":"https://openalex.org/C117011727","wikidata":"https://www.wikidata.org/wiki/Q1278488","display_name":"Bit (key)","level":2,"score":0.6019614934921265},{"id":"https://openalex.org/C28719098","wikidata":"https://www.wikidata.org/wiki/Q44946","display_name":"Point (geometry)","level":2,"score":0.5693848133087158},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.5490116477012634},{"id":"https://openalex.org/C199360897","wikidata":"https://www.wikidata.org/wiki/Q9143","display_name":"Programming language","level":1,"score":0.524160623550415},{"id":"https://openalex.org/C84211073","wikidata":"https://www.wikidata.org/wiki/Q117879","display_name":"Floating point","level":2,"score":0.4209928810596466},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.3864714801311493},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.34753692150115967},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.29931098222732544},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.14535021781921387},{"id":"https://openalex.org/C31258907","wikidata":"https://www.wikidata.org/wiki/Q1301371","display_name":"Computer network","level":1,"score":0.08781862258911133},{"id":"https://openalex.org/C2524010","wikidata":"https://www.wikidata.org/wiki/Q8087","display_name":"Geometry","level":1,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/iceic61013.2024.10457111","is_oa":false,"landing_page_url":"http://dx.doi.org/10.1109/iceic61013.2024.10457111","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2024 International Conference on Electronics, Information, and Communication (ICEIC)","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":39,"referenced_works":["https://openalex.org/W2763421725","https://openalex.org/W2896457183","https://openalex.org/W2970821029","https://openalex.org/W2998617917","https://openalex.org/W3101100041","https://openalex.org/W3194676777","https://openalex.org/W3212496002","https://openalex.org/W4229005866","https://openalex.org/W4288089799","https://openalex.org/W4292760969","https://openalex.org/W4292779060","https://openalex.org/W4298422451","https://openalex.org/W4309591680","https://openalex.org/W4311642023","https://openalex.org/W4322718191","https://openalex.org/W4375868853","https://openalex.org/W4385245566","https://openalex.org/W4386231419","https://openalex.org/W4387114247","https://openalex.org/W4387294240","https://openalex.org/W4387892072","https://openalex.org/W4389519530","https://openalex.org/W6727099177","https://openalex.org/W6739901393","https://openalex.org/W6745245109","https://openalex.org/W6755207826","https://openalex.org/W6767032739","https://openalex.org/W6769627184","https://openalex.org/W6778883912","https://openalex.org/W6785662278","https://openalex.org/W6811340617","https://openalex.org/W6842145542","https://openalex.org/W6847118041","https://openalex.org/W6847478871","https://openalex.org/W6850625674","https://openalex.org/W6856450038","https://openalex.org/W6856696905","https://openalex.org/W6857057920","https://openalex.org/W6857244234"],"related_works":["https://openalex.org/W2411923897","https://openalex.org/W4285347720","https://openalex.org/W4200259850","https://openalex.org/W2333831899","https://openalex.org/W2484894494","https://openalex.org/W2367385042","https://openalex.org/W4381186982","https://openalex.org/W2040781570","https://openalex.org/W4226055750","https://openalex.org/W2079229939"],"abstract_inverted_index":{"Large":[0],"Language":[1],"Models":[2],"(LLMs)":[3],"have":[4],"shown":[5],"remarkable":[6],"success":[7],"in":[8,34,67,112],"various":[9,85],"natural":[10],"language":[11,113],"processing":[12],"tasks.":[13],"However,":[14],"their":[15],"extensive":[16],"parameter":[17],"count":[18],"leads":[19],"to":[20],"significant":[21],"memory":[22],"and":[23,69,90,115],"computational":[24],"demands.":[25],"To":[26],"tackle":[27],"these":[28],"challenges,":[29],"there":[30],"is":[31],"growing":[32],"interest":[33],"employing":[35],"post-training":[36],"quantization":[37],"(PTQ)":[38],"with":[39,108],"reduced-precision":[40],"floating-point":[41,99],"(FP)":[42],"operations.":[43],"Yet,":[44],"the":[45,62,70,84,91],"optimal":[46],"FP":[47],"configuration":[48],"remains":[49],"a":[50,58,80,97,119],"topic":[51],"of":[52,61,83,93,122],"debate.":[53],"Existing":[54],"studies":[55],"often":[56],"overlook":[57],"thorough":[59],"analysis":[60],"diverse":[63],"data":[64,86],"distributions":[65,87],"found":[66],"LLMs":[68,89],"crucial":[71],"design":[72],"choice,":[73],"denormal.":[74],"In":[75],"this":[76],"paper,":[77],"we":[78],"conduct":[79],"comprehensive":[81],"examination":[82],"within":[88],"significance":[92],"denormal":[94],"representation,":[95],"presenting":[96],"mixed-format":[98],"framework.":[100],"Our":[101],"proposed":[102],"framework":[103],"allows":[104],"for":[105],"sub-8-bit":[106],"inference":[107],"minimal":[109],"performance":[110],"degradation":[111],"modeling":[114],"reasoning":[116],"tasks":[117],"across":[118],"broad":[120],"spectrum":[121],"LLMs.":[123]},"counts_by_year":[],"updated_date":"2025-12-25T23:11:45.687758","created_date":"2025-10-10T00:00:00"}