{"id":"https://openalex.org/W7106308190","doi":"https://doi.org/10.1145/3680207.3765259","title":"Elastic On-Device LLM Service","display_name":"Elastic On-Device LLM Service","publication_year":2025,"publication_date":"2025-11-03","ids":{"openalex":"https://openalex.org/W7106308190","doi":"https://doi.org/10.1145/3680207.3765259"},"language":null,"primary_location":{"id":"doi:10.1145/3680207.3765259","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3680207.3765259","pdf_url":"https://dl.acm.org/doi/pdf/10.1145/3680207.3765259","source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 31st Annual International Conference on Mobile Computing and Networking","raw_type":"proceedings-article"},"type":"article","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://dl.acm.org/doi/pdf/10.1145/3680207.3765259","any_repository_has_fulltext":null},"authorships":[{"author_position":"first","author":{"id":null,"display_name":"Wangsong Yin","orcid":"https://orcid.org/0009-0000-6242-4368"},"institutions":[{"id":"https://openalex.org/I20231570","display_name":"Peking University","ror":"https://ror.org/02v51f717","country_code":"CN","type":"education","lineage":["https://openalex.org/I20231570"]}],"countries":["CN"],"is_corresponding":true,"raw_author_name":"Wangsong Yin","raw_affiliation_strings":["Peking University, Beijing, China"],"raw_orcid":"https://orcid.org/0009-0000-6242-4368","affiliations":[{"raw_affiliation_string":"Peking University, Beijing, China","institution_ids":["https://openalex.org/I20231570"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Rongjie Yi","orcid":"https://orcid.org/0009-0003-6040-9596"},"institutions":[{"id":"https://openalex.org/I139759216","display_name":"Beijing University of Posts and Telecommunications","ror":"https://ror.org/04w9fbh59","country_code":"CN","type":"education","lineage":["https://openalex.org/I139759216"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Rongjie Yi","raw_affiliation_strings":["Beijing University of Posts and Telecommunications, Beijing, China"],"raw_orcid":"https://orcid.org/0009-0003-6040-9596","affiliations":[{"raw_affiliation_string":"Beijing University of Posts and Telecommunications, Beijing, China","institution_ids":["https://openalex.org/I139759216"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Daliang Xu","orcid":"https://orcid.org/0000-0002-6775-0688"},"institutions":[{"id":"https://openalex.org/I139759216","display_name":"Beijing University of Posts and Telecommunications","ror":"https://ror.org/04w9fbh59","country_code":"CN","type":"education","lineage":["https://openalex.org/I139759216"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Daliang Xu","raw_affiliation_strings":["Beijing University of Posts and Telecommunications, Beijing, China"],"raw_orcid":"https://orcid.org/0000-0002-6775-0688","affiliations":[{"raw_affiliation_string":"Beijing University of Posts and Telecommunications, Beijing, China","institution_ids":["https://openalex.org/I139759216"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Gang Huang","orcid":"https://orcid.org/0000-0002-4686-3181"},"institutions":[{"id":"https://openalex.org/I20231570","display_name":"Peking University","ror":"https://ror.org/02v51f717","country_code":"CN","type":"education","lineage":["https://openalex.org/I20231570"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Gang Huang","raw_affiliation_strings":["Peking University, Beijing, China"],"raw_orcid":"https://orcid.org/0000-0002-4686-3181","affiliations":[{"raw_affiliation_string":"Peking University, Beijing, China","institution_ids":["https://openalex.org/I20231570"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Mengwei Xu","orcid":"https://orcid.org/0000-0001-6271-6993"},"institutions":[{"id":"https://openalex.org/I139759216","display_name":"Beijing University of Posts and Telecommunications","ror":"https://ror.org/04w9fbh59","country_code":"CN","type":"education","lineage":["https://openalex.org/I139759216"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Mengwei Xu","raw_affiliation_strings":["Beijing University of Posts and Telecommunications, and Beiyou Shenzhen Institute, Beijing, China"],"raw_orcid":"https://orcid.org/0000-0001-6271-6993","affiliations":[{"raw_affiliation_string":"Beijing University of Posts and Telecommunications, and Beiyou Shenzhen Institute, Beijing, China","institution_ids":["https://openalex.org/I139759216"]}]},{"author_position":"last","author":{"id":null,"display_name":"Xuanzhe Liu","orcid":"https://orcid.org/0000-0002-7908-8484"},"institutions":[{"id":"https://openalex.org/I20231570","display_name":"Peking University","ror":"https://ror.org/02v51f717","country_code":"CN","type":"education","lineage":["https://openalex.org/I20231570"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Xuanzhe Liu","raw_affiliation_strings":["Peking University, Beijing, China"],"raw_orcid":"https://orcid.org/0000-0002-7908-8484","affiliations":[{"raw_affiliation_string":"Peking University, Beijing, China","institution_ids":["https://openalex.org/I20231570"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":6,"corresponding_author_ids":[],"corresponding_institution_ids":["https://openalex.org/I20231570"],"apc_list":null,"apc_paid":null,"fwci":1.3517,"has_fulltext":true,"cited_by_count":1,"citation_normalized_percentile":{"value":0.87889225,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":96,"max":98},"biblio":{"volume":null,"issue":null,"first_page":"984","last_page":"999"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T12127","display_name":"Software System Performance and Reliability","score":0.14259999990463257,"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"}},"topics":[{"id":"https://openalex.org/T12127","display_name":"Software System Performance and Reliability","score":0.14259999990463257,"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/T10273","display_name":"IoT and Edge/Fog Computing","score":0.11840000003576279,"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/T14347","display_name":"Big Data and Digital Economy","score":0.11400000005960464,"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"}}],"keywords":[{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.4577000141143799},{"id":"https://openalex.org/keywords/key","display_name":"Key (lock)","score":0.4156000018119812},{"id":"https://openalex.org/keywords/consistency","display_name":"Consistency (knowledge bases)","score":0.39899998903274536},{"id":"https://openalex.org/keywords/satisfiability-modulo-theories","display_name":"Satisfiability modulo theories","score":0.3919000029563904},{"id":"https://openalex.org/keywords/automation","display_name":"Automation","score":0.38589999079704285},{"id":"https://openalex.org/keywords/transformer","display_name":"Transformer","score":0.3801000118255615},{"id":"https://openalex.org/keywords/compiler","display_name":"Compiler","score":0.3781999945640564},{"id":"https://openalex.org/keywords/service","display_name":"Service (business)","score":0.3684000074863434},{"id":"https://openalex.org/keywords/permutation","display_name":"Permutation (music)","score":0.3465000092983246}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7872999906539917},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.4577000141143799},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.4156000018119812},{"id":"https://openalex.org/C120314980","wikidata":"https://www.wikidata.org/wiki/Q180634","display_name":"Distributed computing","level":1,"score":0.4138000011444092},{"id":"https://openalex.org/C2776436953","wikidata":"https://www.wikidata.org/wiki/Q5163215","display_name":"Consistency (knowledge bases)","level":2,"score":0.39899998903274536},{"id":"https://openalex.org/C164155591","wikidata":"https://www.wikidata.org/wiki/Q2067766","display_name":"Satisfiability modulo theories","level":2,"score":0.3919000029563904},{"id":"https://openalex.org/C115901376","wikidata":"https://www.wikidata.org/wiki/Q184199","display_name":"Automation","level":2,"score":0.38589999079704285},{"id":"https://openalex.org/C66322947","wikidata":"https://www.wikidata.org/wiki/Q11658","display_name":"Transformer","level":3,"score":0.3801000118255615},{"id":"https://openalex.org/C169590947","wikidata":"https://www.wikidata.org/wiki/Q47506","display_name":"Compiler","level":2,"score":0.3781999945640564},{"id":"https://openalex.org/C115903868","wikidata":"https://www.wikidata.org/wiki/Q80993","display_name":"Software engineering","level":1,"score":0.3774000108242035},{"id":"https://openalex.org/C2780378061","wikidata":"https://www.wikidata.org/wiki/Q25351891","display_name":"Service (business)","level":2,"score":0.3684000074863434},{"id":"https://openalex.org/C21308566","wikidata":"https://www.wikidata.org/wiki/Q7169365","display_name":"Permutation (music)","level":2,"score":0.3465000092983246},{"id":"https://openalex.org/C2778770139","wikidata":"https://www.wikidata.org/wiki/Q1966904","display_name":"Solver","level":2,"score":0.3434000015258789},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.3292999863624573},{"id":"https://openalex.org/C145644426","wikidata":"https://www.wikidata.org/wiki/Q169411","display_name":"Unified Modeling Language","level":3,"score":0.328900009393692},{"id":"https://openalex.org/C121854251","wikidata":"https://www.wikidata.org/wiki/Q62932","display_name":"Elasticity (physics)","level":2,"score":0.32359999418258667},{"id":"https://openalex.org/C199360897","wikidata":"https://www.wikidata.org/wiki/Q9143","display_name":"Programming language","level":1,"score":0.322299987077713},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.3206000030040741},{"id":"https://openalex.org/C2780451532","wikidata":"https://www.wikidata.org/wiki/Q759676","display_name":"Task (project management)","level":2,"score":0.31130000948905945},{"id":"https://openalex.org/C33676613","wikidata":"https://www.wikidata.org/wiki/Q13415176","display_name":"Dimension (graph theory)","level":2,"score":0.3068999946117401},{"id":"https://openalex.org/C2780165032","wikidata":"https://www.wikidata.org/wiki/Q16869822","display_name":"Energy consumption","level":2,"score":0.30309998989105225},{"id":"https://openalex.org/C136085584","wikidata":"https://www.wikidata.org/wiki/Q910289","display_name":"Overlay","level":2,"score":0.30219998955726624},{"id":"https://openalex.org/C80444323","wikidata":"https://www.wikidata.org/wiki/Q2878974","display_name":"Theoretical computer science","level":1,"score":0.2946000099182129},{"id":"https://openalex.org/C36464697","wikidata":"https://www.wikidata.org/wiki/Q451553","display_name":"Visualization","level":2,"score":0.2944999933242798},{"id":"https://openalex.org/C2777212361","wikidata":"https://www.wikidata.org/wiki/Q5127848","display_name":"Class (philosophy)","level":2,"score":0.2913999855518341},{"id":"https://openalex.org/C186967261","wikidata":"https://www.wikidata.org/wiki/Q5082128","display_name":"Mobile device","level":2,"score":0.29109999537467957},{"id":"https://openalex.org/C2779623668","wikidata":"https://www.wikidata.org/wiki/Q7652842","display_name":"SwIPe","level":2,"score":0.2888999879360199},{"id":"https://openalex.org/C110251889","wikidata":"https://www.wikidata.org/wiki/Q1569697","display_name":"Model checking","level":2,"score":0.28769999742507935},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.2816999852657318},{"id":"https://openalex.org/C2779639559","wikidata":"https://www.wikidata.org/wiki/Q7661178","display_name":"Symbolic execution","level":3,"score":0.27970001101493835},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.27320000529289246},{"id":"https://openalex.org/C79974875","wikidata":"https://www.wikidata.org/wiki/Q483639","display_name":"Cloud computing","level":2,"score":0.2703000009059906},{"id":"https://openalex.org/C527821871","wikidata":"https://www.wikidata.org/wiki/Q228502","display_name":"Access control","level":2,"score":0.2540999948978424},{"id":"https://openalex.org/C113775141","wikidata":"https://www.wikidata.org/wiki/Q428691","display_name":"Computer engineering","level":1,"score":0.25040000677108765}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1145/3680207.3765259","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3680207.3765259","pdf_url":"https://dl.acm.org/doi/pdf/10.1145/3680207.3765259","source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 31st Annual International Conference on Mobile Computing and Networking","raw_type":"proceedings-article"}],"best_oa_location":{"id":"doi:10.1145/3680207.3765259","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3680207.3765259","pdf_url":"https://dl.acm.org/doi/pdf/10.1145/3680207.3765259","source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 31st Annual International Conference on Mobile Computing and Networking","raw_type":"proceedings-article"},"sustainable_development_goals":[{"score":0.4090172350406647,"id":"https://metadata.un.org/sdg/16","display_name":"Peace, Justice and strong institutions"}],"awards":[{"id":"https://openalex.org/G8414340155","display_name":null,"funder_award_id":"62325201","funder_id":"https://openalex.org/F4320321001","funder_display_name":"National Natural Science Foundation of China"}],"funders":[{"id":"https://openalex.org/F4320321001","display_name":"National Natural Science Foundation of China","ror":"https://ror.org/01h0zpd94"}],"has_content":{"pdf":true,"grobid_xml":false},"content_urls":{"pdf":"https://content.openalex.org/works/W7106308190.pdf"},"referenced_works_count":25,"referenced_works":["https://openalex.org/W1566289585","https://openalex.org/W2025713906","https://openalex.org/W2616247523","https://openalex.org/W2890894339","https://openalex.org/W2897268228","https://openalex.org/W2948634622","https://openalex.org/W2979826702","https://openalex.org/W3023960840","https://openalex.org/W3207622241","https://openalex.org/W3210103168","https://openalex.org/W3211149853","https://openalex.org/W4285199320","https://openalex.org/W4380925616","https://openalex.org/W4380928745","https://openalex.org/W4387212784","https://openalex.org/W4387302777","https://openalex.org/W4389519226","https://openalex.org/W4395112660","https://openalex.org/W4398757466","https://openalex.org/W4399121418","https://openalex.org/W4399140776","https://openalex.org/W4402671835","https://openalex.org/W4403333972","https://openalex.org/W4404401018","https://openalex.org/W4404782964"],"related_works":[],"abstract_inverted_index":{"On-device":[0],"Large":[1],"Language":[2],"Models":[3],"(LLMs)":[4],"are":[5],"transforming":[6],"mobile":[7],"AI,":[8],"catalyzing":[9],"applications":[10],"like":[11],"UI":[12],"automation":[13],"without":[14],"privacy":[15],"concerns.":[16],"Nowadays":[17],"the":[18,49,76,79,95,123,127],"common":[19],"practice":[20],"is":[21],"to":[22,51,102,170],"deploy":[23],"a":[24,30,39,83,89,113],"single":[25],"yet":[26],"powerful":[27],"LLM":[28,71,139],"as":[29],"general":[31],"task":[32],"solver":[33],"for":[34],"multiple":[35,142],"requests.":[36],"We":[37,133],"identify":[38],"key":[40],"system":[41],"challenge":[42],"in":[43,99,165],"this":[44],"paradigm:":[45],"current":[46],"LLMs":[47],"lack":[48],"elasticity":[50],"serve":[52],"requests":[53],"that":[54,73],"have":[55],"diversified":[56],"Service-Level":[57],"Objectives":[58],"(SLOs)":[59],"on":[60,141,148,174],"inference":[61],"latency.":[62],"To":[63],"tackle":[64],"this,":[65],"we":[66],"present":[67],"ElastiLM,":[68],"an":[69,136],"on-device":[70,138],"service":[72,140],"elasticizes":[74],"both":[75,149],"model":[77,130],"and":[78,120,125,131,145,153,172,184],"prompt":[80,124],"dimension":[81],"of":[82],"full":[84],"LLM.":[85],"It":[86],"incorporates":[87],"(1)":[88],"one-shot":[90],"neuron-reordering":[91],"method,":[92],"which":[93,118],"leverages":[94],"intrinsic":[96],"permutation":[97],"consistency":[98],"transformer":[100],"models":[101],"generate":[103],"high-quality":[104],"elasticized":[105],"sub-models":[106],"with":[107,176],"minimal":[108],"runtime":[109],"switching":[110,179],"overhead;":[111],"(2)":[112],"dual-head":[114],"tiny":[115],"language":[116],"model,":[117],"efficiently":[119],"effectively":[121],"refines":[122],"orchestrates":[126],"elastification":[128],"between":[129],"prompt.":[132],"implement":[134],"such":[135],"elastic":[137],"COTS":[143],"smartphones,":[144],"evaluate":[146],"ElastiLM":[147,160],"standalone":[150],"NLP/mobile-agent":[151],"datasets":[152],"end-to-end":[154],"synthesized":[155],"traces.":[156],"On":[157],"diverse":[158],"SLOs,":[159],"outperforms":[161],"7":[162],"strong":[163],"baselines":[164],"(absolute)":[166],"accuracy":[167],"by":[168],"up":[169],"14.83%":[171],"10.45%":[173],"average,":[175],"<1%":[177],"TTFT":[178],"overhead,":[180],"on-par":[181],"memory":[182],"consumption":[183],"<100":[185],"offline":[186],"GPU":[187],"hours.":[188]},"counts_by_year":[{"year":2026,"cited_by_count":1}],"updated_date":"2026-03-11T06:11:40.159057","created_date":"2025-11-23T00:00:00"}