{"id":"https://openalex.org/W4400233504","doi":"https://doi.org/10.1109/iscas58744.2024.10558275","title":"Reducing the Energy Dissipation of Large Language Models (LLMs) with Approximate Memories","display_name":"Reducing the Energy Dissipation of Large Language Models (LLMs) with Approximate Memories","publication_year":2024,"publication_date":"2024-05-19","ids":{"openalex":"https://openalex.org/W4400233504","doi":"https://doi.org/10.1109/iscas58744.2024.10558275"},"language":"en","primary_location":{"id":"doi:10.1109/iscas58744.2024.10558275","is_oa":false,"landing_page_url":"https://doi.org/10.1109/iscas58744.2024.10558275","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2024 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/A5101730800","display_name":"Zhen Gao","orcid":"https://orcid.org/0000-0001-5482-4505"},"institutions":[{"id":"https://openalex.org/I162868743","display_name":"Tianjin University","ror":"https://ror.org/012tb2g32","country_code":"CN","type":"education","lineage":["https://openalex.org/I162868743"]}],"countries":["CN"],"is_corresponding":true,"raw_author_name":"Zhen Gao","raw_affiliation_strings":["Tianjin University,School of Electrical and Information Engineering,Tianjin,China"],"affiliations":[{"raw_affiliation_string":"Tianjin University,School of Electrical and Information Engineering,Tianjin,China","institution_ids":["https://openalex.org/I162868743"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5054487153","display_name":"Jie Deng","orcid":"https://orcid.org/0000-0001-9531-8908"},"institutions":[{"id":"https://openalex.org/I162868743","display_name":"Tianjin University","ror":"https://ror.org/012tb2g32","country_code":"CN","type":"education","lineage":["https://openalex.org/I162868743"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Jie Deng","raw_affiliation_strings":["Tianjin University,School of Future Technology,Tianjin,China"],"affiliations":[{"raw_affiliation_string":"Tianjin University,School of Future Technology,Tianjin,China","institution_ids":["https://openalex.org/I162868743"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5080322790","display_name":"Pedro Reviriego","orcid":"https://orcid.org/0000-0003-2540-5234"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Pedro Reviriego","raw_affiliation_strings":["Universidad Politécnica de Madrid,Departamento de Ingeniería de Sistemas Telemáticos,Madrid"],"affiliations":[{"raw_affiliation_string":"Universidad Politécnica de Madrid,Departamento de Ingeniería de Sistemas Telemáticos,Madrid","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100417119","display_name":"Shanshan Liu","orcid":"https://orcid.org/0000-0001-6226-2880"},"institutions":[{"id":"https://openalex.org/I150229711","display_name":"University of Electronic Science and Technology of China","ror":"https://ror.org/04qr3zq92","country_code":"CN","type":"education","lineage":["https://openalex.org/I150229711"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Shanshan Liu","raw_affiliation_strings":["University of Electronic Science and Technology of China,School of Information and Communication Engineering,Chengdu,China"],"affiliations":[{"raw_affiliation_string":"University of Electronic Science and Technology of China,School of Information and Communication Engineering,Chengdu,China","institution_ids":["https://openalex.org/I150229711"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5001979328","display_name":"Fabrizio Lombardi","orcid":"https://orcid.org/0000-0003-3152-3245"},"institutions":[{"id":"https://openalex.org/I12912129","display_name":"Northeastern University","ror":"https://ror.org/04t5xt781","country_code":"US","type":"education","lineage":["https://openalex.org/I12912129"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Fabrizio Lombardi","raw_affiliation_strings":["Northeastern University,Department of Electrical and Computer Engineering,Boston,MA,USA"],"affiliations":[{"raw_affiliation_string":"Northeastern University,Department of Electrical and Computer Engineering,Boston,MA,USA","institution_ids":["https://openalex.org/I12912129"]}]}],"institutions":[],"countries_distinct_count":2,"institutions_distinct_count":5,"corresponding_author_ids":["https://openalex.org/A5101730800"],"corresponding_institution_ids":["https://openalex.org/I162868743"],"apc_list":null,"apc_paid":null,"fwci":1.4504,"has_fulltext":false,"cited_by_count":4,"citation_normalized_percentile":{"value":0.84201569,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":94,"max":96},"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/T10028","display_name":"Topic Modeling","score":0.8580999970436096,"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/T10028","display_name":"Topic Modeling","score":0.8580999970436096,"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/dissipation","display_name":"Dissipation","score":0.7920025587081909},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.5102963447570801},{"id":"https://openalex.org/keywords/thermal-management-of-electronic-devices-and-systems","display_name":"Thermal management of electronic devices and systems","score":0.4809950590133667},{"id":"https://openalex.org/keywords/energy","display_name":"Energy (signal processing)","score":0.4736016094684601},{"id":"https://openalex.org/keywords/physics","display_name":"Physics","score":0.2016592025756836},{"id":"https://openalex.org/keywords/engineering","display_name":"Engineering","score":0.13863351941108704},{"id":"https://openalex.org/keywords/mechanical-engineering","display_name":"Mechanical engineering","score":0.09959849715232849},{"id":"https://openalex.org/keywords/thermodynamics","display_name":"Thermodynamics","score":0.07976385951042175}],"concepts":[{"id":"https://openalex.org/C135402231","wikidata":"https://www.wikidata.org/wiki/Q898440","display_name":"Dissipation","level":2,"score":0.7920025587081909},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.5102963447570801},{"id":"https://openalex.org/C114834414","wikidata":"https://www.wikidata.org/wiki/Q15477170","display_name":"Thermal management of electronic devices and systems","level":2,"score":0.4809950590133667},{"id":"https://openalex.org/C186370098","wikidata":"https://www.wikidata.org/wiki/Q442787","display_name":"Energy (signal processing)","level":2,"score":0.4736016094684601},{"id":"https://openalex.org/C121332964","wikidata":"https://www.wikidata.org/wiki/Q413","display_name":"Physics","level":0,"score":0.2016592025756836},{"id":"https://openalex.org/C127413603","wikidata":"https://www.wikidata.org/wiki/Q11023","display_name":"Engineering","level":0,"score":0.13863351941108704},{"id":"https://openalex.org/C78519656","wikidata":"https://www.wikidata.org/wiki/Q101333","display_name":"Mechanical engineering","level":1,"score":0.09959849715232849},{"id":"https://openalex.org/C97355855","wikidata":"https://www.wikidata.org/wiki/Q11473","display_name":"Thermodynamics","level":1,"score":0.07976385951042175},{"id":"https://openalex.org/C62520636","wikidata":"https://www.wikidata.org/wiki/Q944","display_name":"Quantum mechanics","level":1,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/iscas58744.2024.10558275","is_oa":false,"landing_page_url":"https://doi.org/10.1109/iscas58744.2024.10558275","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2024 IEEE International Symposium on Circuits and Systems (ISCAS)","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[{"score":0.8899999856948853,"id":"https://metadata.un.org/sdg/7","display_name":"Affordable and clean energy"}],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":25,"referenced_works":["https://openalex.org/W2092582556","https://openalex.org/W2112704250","https://openalex.org/W2149743155","https://openalex.org/W2193198328","https://openalex.org/W2516832665","https://openalex.org/W2753050160","https://openalex.org/W2978017171","https://openalex.org/W2996428491","https://openalex.org/W3000623753","https://openalex.org/W3009178428","https://openalex.org/W3197266043","https://openalex.org/W4292779060","https://openalex.org/W4312607490","https://openalex.org/W4377371819","https://openalex.org/W4385890089","https://openalex.org/W4386273009","https://openalex.org/W4387293242","https://openalex.org/W6755207826","https://openalex.org/W6768021236","https://openalex.org/W6768851824","https://openalex.org/W6778883912","https://openalex.org/W6852962002","https://openalex.org/W6855616086","https://openalex.org/W6855780230","https://openalex.org/W6856823467"],"related_works":["https://openalex.org/W4391375266","https://openalex.org/W2748952813","https://openalex.org/W2347486132","https://openalex.org/W2316789606","https://openalex.org/W2350340797","https://openalex.org/W4293224283","https://openalex.org/W2950501077","https://openalex.org/W2368601041","https://openalex.org/W2066200948","https://openalex.org/W2140614965"],"abstract_inverted_index":{"Large":[0],"language":[1],"models":[2],"(LLMs)":[3],"have":[4,74,164],"shown":[5],"impressive":[6],"performance":[7,213],"in":[8,144,186,218],"a":[9,173,207,215],"wide":[10],"range":[11],"of":[12,34,46,95,113,122,141,146,153,172,203],"tasks":[13],"such":[14],"as":[15,29,92],"answering":[16],"questions":[17],"or":[18],"summarizing":[19],"text.":[20],"However,":[21],"running":[22],"LLMs":[23,44,156],"on":[24,155,168,210],"edge":[25],"devices":[26],"is":[27,49,84],"challenging":[28],"they":[30],"require":[31],"large":[32],"amounts":[33],"energy":[35,104,136,219],"due":[36],"to":[37,51,64,77,87,102,109,125],"their":[38,89],"memory":[39,48,90,114,148],"and":[40,79,214],"computation":[41],"needs.":[42],"In":[43,67,99],"most":[45],"the":[47,53,65,68,93,111,120,127,135,139,147,151,201,211],"needed":[50],"store":[52,126],"model":[54],"parameters":[55,96],"which":[56],"number":[57,94],"keeps":[58],"increasing":[59],"from":[60,179],"one":[61],"LLM":[62,128,212],"generation":[63],"next.":[66],"last":[69],"several":[70],"years,":[71],"significant":[72,216],"efforts":[73],"been":[75],"made":[76],"compress":[78],"prune":[80],"parameters,":[81],"but":[82],"this":[83,100],"not":[85],"enough":[86],"reduce":[88,103,110,134],"needs":[91],"grows":[97],"exponentially.":[98],"work,":[101],"dissipation,":[105],"rather":[106],"than":[107],"trying":[108],"amount":[112],"used":[115],"by":[116],"LLMs,":[117],"we":[118,163],"study":[119],"use":[121,202],"approximate":[123,204],"memories":[124,131,205],"parameters.":[129],"Approximate":[130],"can":[132],"significantly":[133],"dissipation":[137],"at":[138,193],"cost":[140],"introducing":[142],"errors":[143,154],"some":[145,187],"bits.":[149],"Therefore,":[150],"impact":[152,209],"must":[157],"be":[158],"understood.":[159],"To":[160],"that":[161,185],"end,":[162],"performed":[165],"error":[166,196],"injection":[167],"different":[169],"compressed":[170,189],"versions":[171],"classic":[174],"LLM:":[175],"Bidirectional":[176],"Encoder":[177],"Representations":[178],"Transformers":[180],"(BERT).":[181],"The":[182],"results":[183],"show":[184],"cases":[188],"BERTs":[190],"operate":[191],"reliably":[192],"high":[194],"bit":[195],"rates.":[197],"This":[198],"makes":[199],"possible":[200],"with":[206],"negligible":[208],"reduction":[217],"dissipation.":[220]},"counts_by_year":[{"year":2025,"cited_by_count":2},{"year":2024,"cited_by_count":2}],"updated_date":"2025-12-19T19:40:27.379048","created_date":"2025-10-10T00:00:00"}