{"id":"https://openalex.org/W4416963822","doi":"https://doi.org/10.1109/islped65674.2025.11261772","title":"QuAKE: Speeding up Model Inference Using Quick and Approximate Kernels for Exponential Non-Linearities","display_name":"QuAKE: Speeding up Model Inference Using Quick and Approximate Kernels for Exponential Non-Linearities","publication_year":2025,"publication_date":"2025-08-06","ids":{"openalex":"https://openalex.org/W4416963822","doi":"https://doi.org/10.1109/islped65674.2025.11261772"},"language":null,"primary_location":{"id":"doi:10.1109/islped65674.2025.11261772","is_oa":false,"landing_page_url":"https://doi.org/10.1109/islped65674.2025.11261772","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2025 IEEE/ACM International Symposium on Low Power Electronics and Design (ISLPED)","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/A5003531741","display_name":"Sai Kiran Narayanaswami","orcid":null},"institutions":[{"id":"https://openalex.org/I24676775","display_name":"Indian Institute of Technology Madras","ror":"https://ror.org/03v0r5n49","country_code":"IN","type":"facility","lineage":["https://openalex.org/I24676775"]}],"countries":["IN"],"is_corresponding":true,"raw_author_name":"Sai Kiran Narayanaswami","raw_affiliation_strings":["Indian Institute of Technology,Centre for Responsible AI,Madras"],"affiliations":[{"raw_affiliation_string":"Indian Institute of Technology,Centre for Responsible AI,Madras","institution_ids":["https://openalex.org/I24676775"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5011164681","display_name":"Gopalakrishnan Srinivasan","orcid":"https://orcid.org/0000-0003-2015-8545"},"institutions":[{"id":"https://openalex.org/I4210151956","display_name":"Robert Bosch (India)","ror":"https://ror.org/04my8ty22","country_code":"IN","type":"company","lineage":["https://openalex.org/I4210151956","https://openalex.org/I889804353"]}],"countries":["IN"],"is_corresponding":false,"raw_author_name":"Gopalakrishnan Srinivasan","raw_affiliation_strings":["Indian Institute of Technology,Robert Bosch Centre for Data Science and AI,Dept. of Computer Science and Engineering,Madras"],"affiliations":[{"raw_affiliation_string":"Indian Institute of Technology,Robert Bosch Centre for Data Science and AI,Dept. of Computer Science and Engineering,Madras","institution_ids":["https://openalex.org/I4210151956"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5009374923","display_name":"Balaraman Ravindran","orcid":"https://orcid.org/0000-0002-5364-7639"},"institutions":[{"id":"https://openalex.org/I24676775","display_name":"Indian Institute of Technology Madras","ror":"https://ror.org/03v0r5n49","country_code":"IN","type":"facility","lineage":["https://openalex.org/I24676775"]}],"countries":["IN"],"is_corresponding":false,"raw_author_name":"Balaraman Ravindran","raw_affiliation_strings":["Indian Institute of Technology,Wadhwani School of Data Science and AI, Centre for Responsible AI,Madras"],"affiliations":[{"raw_affiliation_string":"Indian Institute of Technology,Wadhwani School of Data Science and AI, Centre for Responsible AI,Madras","institution_ids":["https://openalex.org/I24676775"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":3,"corresponding_author_ids":["https://openalex.org/A5003531741"],"corresponding_institution_ids":["https://openalex.org/I24676775"],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":{"value":0.3947192,"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":"7"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T11697","display_name":"Numerical Methods and Algorithms","score":0.4277999997138977,"subfield":{"id":"https://openalex.org/subfields/1703","display_name":"Computational Theory and Mathematics"},"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/T11697","display_name":"Numerical Methods and Algorithms","score":0.4277999997138977,"subfield":{"id":"https://openalex.org/subfields/1703","display_name":"Computational Theory and Mathematics"},"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/T11206","display_name":"Model Reduction and Neural Networks","score":0.07050000131130219,"subfield":{"id":"https://openalex.org/subfields/3109","display_name":"Statistical and Nonlinear Physics"},"field":{"id":"https://openalex.org/fields/31","display_name":"Physics and Astronomy"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T10036","display_name":"Advanced Neural Network Applications","score":0.05820000171661377,"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"}}],"keywords":[{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.7226999998092651},{"id":"https://openalex.org/keywords/leverage","display_name":"Leverage (statistics)","score":0.7021999955177307},{"id":"https://openalex.org/keywords/exponential-function","display_name":"Exponential function","score":0.6687999963760376},{"id":"https://openalex.org/keywords/computation","display_name":"Computation","score":0.5442000031471252},{"id":"https://openalex.org/keywords/approximate-inference","display_name":"Approximate inference","score":0.5019999742507935},{"id":"https://openalex.org/keywords/key","display_name":"Key (lock)","score":0.38019999861717224},{"id":"https://openalex.org/keywords/quake","display_name":"Quake (natural phenomenon)","score":0.375900000333786},{"id":"https://openalex.org/keywords/exponential-growth","display_name":"Exponential growth","score":0.35019999742507935}],"concepts":[{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.7226999998092651},{"id":"https://openalex.org/C153083717","wikidata":"https://www.wikidata.org/wiki/Q6535263","display_name":"Leverage (statistics)","level":2,"score":0.7021999955177307},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.689300000667572},{"id":"https://openalex.org/C151376022","wikidata":"https://www.wikidata.org/wiki/Q168698","display_name":"Exponential function","level":2,"score":0.6687999963760376},{"id":"https://openalex.org/C45374587","wikidata":"https://www.wikidata.org/wiki/Q12525525","display_name":"Computation","level":2,"score":0.5442000031471252},{"id":"https://openalex.org/C2777472644","wikidata":"https://www.wikidata.org/wiki/Q16968992","display_name":"Approximate inference","level":3,"score":0.5019999742507935},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.400299996137619},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.3837999999523163},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.38019999861717224},{"id":"https://openalex.org/C40160016","wikidata":"https://www.wikidata.org/wiki/Q2692116","display_name":"Quake (natural phenomenon)","level":2,"score":0.375900000333786},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.36169999837875366},{"id":"https://openalex.org/C75235859","wikidata":"https://www.wikidata.org/wiki/Q582659","display_name":"Exponential growth","level":2,"score":0.35019999742507935},{"id":"https://openalex.org/C134261354","wikidata":"https://www.wikidata.org/wiki/Q938438","display_name":"Statistical inference","level":2,"score":0.33889999985694885},{"id":"https://openalex.org/C80444323","wikidata":"https://www.wikidata.org/wiki/Q2878974","display_name":"Theoretical computer science","level":1,"score":0.32280001044273376},{"id":"https://openalex.org/C14036430","wikidata":"https://www.wikidata.org/wiki/Q3736076","display_name":"Function (biology)","level":2,"score":0.3176000118255615},{"id":"https://openalex.org/C49016629","wikidata":"https://www.wikidata.org/wiki/Q3075235","display_name":"Double exponential function","level":3,"score":0.31679999828338623},{"id":"https://openalex.org/C55974624","wikidata":"https://www.wikidata.org/wiki/Q1188504","display_name":"Exponential family","level":2,"score":0.3122999966144562},{"id":"https://openalex.org/C68339613","wikidata":"https://www.wikidata.org/wiki/Q1549489","display_name":"Speedup","level":2,"score":0.3061999976634979},{"id":"https://openalex.org/C126255220","wikidata":"https://www.wikidata.org/wiki/Q141495","display_name":"Mathematical optimization","level":1,"score":0.29919999837875366},{"id":"https://openalex.org/C114289077","wikidata":"https://www.wikidata.org/wiki/Q3284399","display_name":"Statistical model","level":2,"score":0.27559998631477356},{"id":"https://openalex.org/C88871306","wikidata":"https://www.wikidata.org/wiki/Q7208287","display_name":"Point process","level":2,"score":0.273499995470047},{"id":"https://openalex.org/C28719098","wikidata":"https://www.wikidata.org/wiki/Q44946","display_name":"Point (geometry)","level":2,"score":0.2669999897480011},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.26499998569488525}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/islped65674.2025.11261772","is_oa":false,"landing_page_url":"https://doi.org/10.1109/islped65674.2025.11261772","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2025 IEEE/ACM International Symposium on Low Power Electronics and Design (ISLPED)","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":19,"referenced_works":["https://openalex.org/W1494198834","https://openalex.org/W2108598243","https://openalex.org/W2111013824","https://openalex.org/W2137269967","https://openalex.org/W2137983211","https://openalex.org/W2800818619","https://openalex.org/W2902032143","https://openalex.org/W2946609015","https://openalex.org/W3046911311","https://openalex.org/W3138516171","https://openalex.org/W3211525823","https://openalex.org/W4206087890","https://openalex.org/W4283706230","https://openalex.org/W4293025109","https://openalex.org/W4312443924","https://openalex.org/W4385080022","https://openalex.org/W4385245566","https://openalex.org/W4390970343","https://openalex.org/W4402349219"],"related_works":[],"abstract_inverted_index":{"As":[0],"machine":[1],"learning":[2],"gets":[3],"deployed":[4],"more":[5,7],"and":[6,9,101,114,119,121,125,134,143],"widely,":[8],"model":[10,19,30,132,138],"sizes":[11],"continue":[12],"to":[13,69,154,161],"grow,":[14],"improving":[15],"computational":[16],"efficiency":[17,89],"during":[18],"inference":[20,39,109],"has":[21],"become":[22],"a":[23,34,55,129],"key":[24],"challenge.":[25],"In":[26,49],"many":[27],"commonly":[28,93],"used":[29,94],"architectures,":[31],"including":[32],"Transformers,":[33],"significant":[35],"portion":[36],"of":[37,43,57,64,90,131,137,164],"the":[38,72,88,102],"computation":[40],"is":[41],"comprised":[42],"exponential":[44,73,95],"non-linearities":[45,96],"such":[46,97],"as":[47,98],"Softmax.":[48],"this":[50],"work,":[51],"we":[52],"develop":[53],"QuAKE,":[54],"collection":[56],"novel":[58],"operators":[59,151],"that":[60,86,149],"leverage":[61],"certain":[62],"properties":[63],"IEEE-754":[65],"floating":[66],"point":[67],"representations":[68],"quickly":[70],"approximate":[71],"function":[74],"without":[75],"requiring":[76],"specialized":[77],"hardware,":[78],"extra":[79],"memory,":[80],"or":[81],"precomputation.":[82],"We":[83],"propose":[84],"optimizations":[85],"enhance":[87],"QuAKE":[91,150],"in":[92],"Softmax,":[99],"GELU,":[100],"Logistic":[103],"function.":[104],"Our":[105],"benchmarks":[106],"demonstrate":[107],"substantial":[108],"speed":[110,157],"improvements":[111],"between":[112],"10%":[113],"35%":[115],"on":[116,123,140,166],"server":[117],"CPUs,":[118],"5%":[120],"45%":[122],"embedded":[124],"mobile-scale":[126],"CPUs":[127],"for":[128],"variety":[130],"architectures":[133],"sizes.":[135],"Evaluations":[136],"performance":[139,165],"standard":[141],"datasets":[142],"tasks":[144],"from":[145],"various":[146],"domains":[147],"show":[148],"are":[152],"able":[153],"provide":[155],"sizable":[156],"benefits":[158],"with":[159],"little":[160],"no":[162],"loss":[163],"downstream":[167],"tasks.":[168]},"counts_by_year":[],"updated_date":"2026-03-07T16:01:11.037858","created_date":"2025-12-03T00:00:00"}