{"id":"https://openalex.org/W4412945503","doi":"https://doi.org/10.18653/v1/2025.acl-long.482","title":"A Drop-In Solution for On-the-Fly Adaptation of Speculative Decoding in Large Language Models","display_name":"A Drop-In Solution for On-the-Fly Adaptation of Speculative Decoding in Large Language Models","publication_year":2025,"publication_date":"2025-01-01","ids":{"openalex":"https://openalex.org/W4412945503","doi":"https://doi.org/10.18653/v1/2025.acl-long.482"},"language":"en","primary_location":{"id":"doi:10.18653/v1/2025.acl-long.482","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.acl-long.482","pdf_url":"https://aclanthology.org/2025.acl-long.482.pdf","source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)","raw_type":"proceedings-article"},"type":"conference-paper","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://aclanthology.org/2025.acl-long.482.pdf","any_repository_has_fulltext":null},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5047469209","display_name":"Jiesong Liu","orcid":"https://orcid.org/0000-0002-8311-020X"},"institutions":[{"id":"https://openalex.org/I137902535","display_name":"North Carolina State University","ror":"https://ror.org/04tj63d06","country_code":"US","type":"education","lineage":["https://openalex.org/I137902535"]},{"id":"https://openalex.org/I35777872","display_name":"North Carolina Agricultural and Technical State University","ror":"https://ror.org/02aze4h65","country_code":"US","type":"education","lineage":["https://openalex.org/I35777872"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Jiesong Liu","raw_affiliation_strings":["Department of Computer Science , North Carolina State University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Department of Computer Science , North Carolina State University","institution_ids":["https://openalex.org/I137902535","https://openalex.org/I35777872"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100663774","display_name":"Brian Park","orcid":"https://orcid.org/0009-0004-8795-5949"},"institutions":[{"id":"https://openalex.org/I137902535","display_name":"North Carolina State University","ror":"https://ror.org/04tj63d06","country_code":"US","type":"education","lineage":["https://openalex.org/I137902535"]},{"id":"https://openalex.org/I35777872","display_name":"North Carolina Agricultural and Technical State University","ror":"https://ror.org/02aze4h65","country_code":"US","type":"education","lineage":["https://openalex.org/I35777872"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Brian Park","raw_affiliation_strings":["Department of Computer Science , North Carolina State University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Department of Computer Science , North Carolina State University","institution_ids":["https://openalex.org/I137902535","https://openalex.org/I35777872"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5100624451","display_name":"Xipeng Shen","orcid":"https://orcid.org/0000-0003-3599-8010"},"institutions":[{"id":"https://openalex.org/I137902535","display_name":"North Carolina State University","ror":"https://ror.org/04tj63d06","country_code":"US","type":"education","lineage":["https://openalex.org/I137902535"]},{"id":"https://openalex.org/I35777872","display_name":"North Carolina Agricultural and Technical State University","ror":"https://ror.org/02aze4h65","country_code":"US","type":"education","lineage":["https://openalex.org/I35777872"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Xipeng Shen","raw_affiliation_strings":["Department of Computer Science , North Carolina State University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Department of Computer Science , North Carolina State University","institution_ids":["https://openalex.org/I137902535","https://openalex.org/I35777872"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":2,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":true,"cited_by_count":0,"citation_normalized_percentile":{"value":0.15344255,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"9778","last_page":"9794"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10028","display_name":"Topic Modeling","score":0.9901999831199646,"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":0.9901999831199646,"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"}},{"id":"https://openalex.org/T10181","display_name":"Natural Language Processing Techniques","score":0.9876999855041504,"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/decoding-methods","display_name":"Decoding methods","score":0.7347995042800903},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.6891538500785828},{"id":"https://openalex.org/keywords/adaptation","display_name":"Adaptation (eye)","score":0.5513408780097961},{"id":"https://openalex.org/keywords/drop","display_name":"Drop (telecommunication)","score":0.535434365272522},{"id":"https://openalex.org/keywords/on-the-fly","display_name":"On the fly","score":0.4656762480735779},{"id":"https://openalex.org/keywords/parallel-computing","display_name":"Parallel computing","score":0.3572345972061157},{"id":"https://openalex.org/keywords/algorithm","display_name":"Algorithm","score":0.17121392488479614},{"id":"https://openalex.org/keywords/operating-system","display_name":"Operating system","score":0.16022074222564697},{"id":"https://openalex.org/keywords/telecommunications","display_name":"Telecommunications","score":0.13609111309051514},{"id":"https://openalex.org/keywords/psychology","display_name":"Psychology","score":0.10067188739776611}],"concepts":[{"id":"https://openalex.org/C57273362","wikidata":"https://www.wikidata.org/wiki/Q576722","display_name":"Decoding methods","level":2,"score":0.7347995042800903},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6891538500785828},{"id":"https://openalex.org/C139807058","wikidata":"https://www.wikidata.org/wiki/Q352374","display_name":"Adaptation (eye)","level":2,"score":0.5513408780097961},{"id":"https://openalex.org/C2781345722","wikidata":"https://www.wikidata.org/wiki/Q5308388","display_name":"Drop (telecommunication)","level":2,"score":0.535434365272522},{"id":"https://openalex.org/C2781020372","wikidata":"https://www.wikidata.org/wiki/Q533093","display_name":"On the fly","level":2,"score":0.4656762480735779},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.3572345972061157},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.17121392488479614},{"id":"https://openalex.org/C111919701","wikidata":"https://www.wikidata.org/wiki/Q9135","display_name":"Operating system","level":1,"score":0.16022074222564697},{"id":"https://openalex.org/C76155785","wikidata":"https://www.wikidata.org/wiki/Q418","display_name":"Telecommunications","level":1,"score":0.13609111309051514},{"id":"https://openalex.org/C15744967","wikidata":"https://www.wikidata.org/wiki/Q9418","display_name":"Psychology","level":0,"score":0.10067188739776611},{"id":"https://openalex.org/C169760540","wikidata":"https://www.wikidata.org/wiki/Q207011","display_name":"Neuroscience","level":1,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.18653/v1/2025.acl-long.482","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.acl-long.482","pdf_url":"https://aclanthology.org/2025.acl-long.482.pdf","source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)","raw_type":"proceedings-article"}],"best_oa_location":{"id":"doi:10.18653/v1/2025.acl-long.482","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.acl-long.482","pdf_url":"https://aclanthology.org/2025.acl-long.482.pdf","source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)","raw_type":"proceedings-article"},"sustainable_development_goals":[],"awards":[{"id":"https://openalex.org/G2350003802","display_name":"Collaborative Research: CSR: Medium: Scaling Secure Serverless Computing on Heterogeneous Datacenters","funder_award_id":"2312207","funder_id":"https://openalex.org/F4320306076","funder_display_name":"National Science Foundation"},{"id":"https://openalex.org/G5028539974","display_name":null,"funder_award_id":"CNS-2312207","funder_id":"https://openalex.org/F4320306076","funder_display_name":"National Science Foundation"},{"id":"https://openalex.org/G5852758417","display_name":null,"funder_award_id":"1R01HD108473","funder_id":"https://openalex.org/F4320332161","funder_display_name":"National Institutes of Health"}],"funders":[{"id":"https://openalex.org/F4320306076","display_name":"National Science Foundation","ror":"https://ror.org/021nxhr62"},{"id":"https://openalex.org/F4320332161","display_name":"National Institutes of Health","ror":"https://ror.org/01cwqze88"}],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4412945503.pdf","grobid_xml":"https://content.openalex.org/works/W4412945503.grobid-xml"},"referenced_works_count":1,"referenced_works":["https://openalex.org/W4391335474"],"related_works":["https://openalex.org/W2161474341","https://openalex.org/W2997567050","https://openalex.org/W4302615923","https://openalex.org/W3203142394","https://openalex.org/W2351061015","https://openalex.org/W4220731478","https://openalex.org/W1974101135","https://openalex.org/W2017509870","https://openalex.org/W4251141768","https://openalex.org/W4360952157"],"abstract_inverted_index":{"Large":[0],"Language":[1],"Models":[2],"(LLMs)":[3],"are":[4],"cuttingedge":[5],"generative":[6],"AI":[7],"models":[8],"built":[9],"on":[10],"transformer":[11],"architecture,":[12],"which":[13,37],"tend":[14],"to":[15,27,88,91,131,157],"be":[16],"highly":[17],"memoryintensive":[18],"when":[19],"performing":[20],"real-time":[21],"inference.Various":[22],"strategies":[23],"have":[24],"been":[25],"developed":[26],"enhance":[28],"the":[29,64,69,73,80,92,99,102,104,109,112,129,133,153,161,167],"end-to-end":[30],"inference":[31,50],"speed":[32],"for":[33,49,78,138],"LLMs,":[34],"one":[35],"of":[36,82,101,111,121,135],"is":[38,76,86],"speculative":[39,83,122,136,163],"decoding.This":[40],"technique":[41],"involves":[42],"running":[43],"a":[44,52,124,141],"smaller":[45],"LLM":[46,66,139],"(draft":[47],"model)":[48],"over":[51,160],"defined":[53],"window":[54],"size,":[55],"denoted":[56],"as":[57],",":[58],"while":[59],"simultaneously":[60],"being":[61],"validated":[62],"by":[63],"larger":[65],"(target":[67],"model).Choosing":[68],"optimal":[70],"value":[71],"and":[72,108,114,165],"draft":[74],"model":[75],"essential":[77],"unlocking":[79],"potential":[81],"decoding.But":[84],"it":[85,144],"difficult":[87],"do":[89],"due":[90],"complicated":[93],"influence":[94],"from":[95],"various":[96],"factors,":[97],"including":[98],"nature":[100],"task,":[103],"hardware":[105],"in":[106],"use,":[107],"combination":[110],"large":[113],"small":[115],"models.This":[116],"paper":[117],"introduces":[118],"on-the-fly":[119],"adaption":[120],"decoding,":[123,164],"solution":[125,154],"that":[126,152],"dynamically":[127],"adapts":[128],"choices":[130],"maximize":[132],"efficiency":[134],"decoding":[137],"inferences.As":[140],"drop-in":[142],"solution,":[143],"needs":[145],"no":[146],"offline":[147],"benchmarking":[148],"or":[149],"training.Experiments":[150],"show":[151],"can":[155],"lead":[156],"3.55-16.48%speed":[158],"improvement":[159],"standard":[162],"1.2-3.4over":[166],"default":[168],"LLMs.":[169]},"counts_by_year":[],"updated_date":"2026-07-29T14:22:42.915294","created_date":"2025-10-10T00:00:00"}