{"id":"https://openalex.org/W4410294680","doi":"https://doi.org/10.26599/bdma.2024.9020074","title":"ResDecode: Accelerating Large Language Models Inference via Residual Decoding Heads","display_name":"ResDecode: Accelerating Large Language Models Inference via Residual Decoding Heads","publication_year":2025,"publication_date":"2025-05-12","ids":{"openalex":"https://openalex.org/W4410294680","doi":"https://doi.org/10.26599/bdma.2024.9020074"},"language":"en","primary_location":{"id":"doi:10.26599/bdma.2024.9020074","is_oa":true,"landing_page_url":"https://doi.org/10.26599/bdma.2024.9020074","pdf_url":null,"source":{"id":"https://openalex.org/S4210209060","display_name":"Big Data Mining and Analytics","issn_l":"2096-0654","issn":["2096-0654","2097-406X"],"is_oa":true,"is_in_doaj":true,"is_core":true,"host_organization":"https://openalex.org/P4310311901","host_organization_name":"Tsinghua University Press","host_organization_lineage":["https://openalex.org/P4310311901"],"host_organization_lineage_names":["Tsinghua University Press"],"type":"journal"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Big Data Mining and Analytics","raw_type":"journal-article"},"type":"article","indexed_in":["crossref","doaj"],"open_access":{"is_oa":true,"oa_status":"diamond","oa_url":"https://doi.org/10.26599/bdma.2024.9020074","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5101403806","display_name":"Ziqian Zeng","orcid":"https://orcid.org/0000-0003-0060-7956"},"institutions":[{"id":"https://openalex.org/I90610280","display_name":"South China University of Technology","ror":"https://ror.org/0530pts50","country_code":"CN","type":"education","lineage":["https://openalex.org/I90610280"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Ziqian Zeng","raw_affiliation_strings":["Shien Ming Wu School of Intelligent Engineering, South China University of Technology,Guangzhou,China,511442"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Shien Ming Wu School of Intelligent Engineering, South China University of Technology,Guangzhou,China,511442","institution_ids":["https://openalex.org/I90610280"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5102415308","display_name":"Jiahong Yu","orcid":null},"institutions":[{"id":"https://openalex.org/I90610280","display_name":"South China University of Technology","ror":"https://ror.org/0530pts50","country_code":"CN","type":"education","lineage":["https://openalex.org/I90610280"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Jiahong Yu","raw_affiliation_strings":["School of Computer Science and Engineering, South China University of Technology,Guangzhou,China,510006"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"School of Computer Science and Engineering, South China University of Technology,Guangzhou,China,510006","institution_ids":["https://openalex.org/I90610280"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5111136901","display_name":"Qianshi Pang","orcid":null},"institutions":[{"id":"https://openalex.org/I90610280","display_name":"South China University of Technology","ror":"https://ror.org/0530pts50","country_code":"CN","type":"education","lineage":["https://openalex.org/I90610280"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Qianshi Pang","raw_affiliation_strings":["School of Computer Science and Engineering, South China University of Technology,Guangzhou,China,510006"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"School of Computer Science and Engineering, South China University of Technology,Guangzhou,China,510006","institution_ids":["https://openalex.org/I90610280"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100413085","display_name":"Zihao Wang","orcid":"https://orcid.org/0000-0002-3919-0396"},"institutions":[{"id":"https://openalex.org/I200769079","display_name":"Hong Kong University of Science and Technology","ror":"https://ror.org/00q4vv597","country_code":"HK","type":"education","lineage":["https://openalex.org/I200769079"]},{"id":"https://openalex.org/I889458895","display_name":"University of Hong Kong","ror":"https://ror.org/02zhqgq86","country_code":"HK","type":"education","lineage":["https://openalex.org/I889458895"]}],"countries":["HK"],"is_corresponding":false,"raw_author_name":"Zihao Wang","raw_affiliation_strings":["School of Engineering, The Hong Kong University of Science and Technology,Department of Computer Science and Engineering,Hong Kong,China,999077"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"School of Engineering, The Hong Kong University of Science and Technology,Department of Computer Science and Engineering,Hong Kong,China,999077","institution_ids":["https://openalex.org/I200769079","https://openalex.org/I889458895"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5061256037","display_name":"Huiping Zhuang","orcid":"https://orcid.org/0000-0002-4612-5445"},"institutions":[{"id":"https://openalex.org/I90610280","display_name":"South China University of Technology","ror":"https://ror.org/0530pts50","country_code":"CN","type":"education","lineage":["https://openalex.org/I90610280"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Huiping Zhuang","raw_affiliation_strings":["Shien Ming Wu School of Intelligent Engineering, South China University of Technology,Guangzhou,China,511442"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Shien Ming Wu School of Intelligent Engineering, South China University of Technology,Guangzhou,China,511442","institution_ids":["https://openalex.org/I90610280"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101814062","display_name":"Yu Fan","orcid":"https://orcid.org/0009-0001-2189-351X"},"institutions":[{"id":"https://openalex.org/I2250955327","display_name":"Huawei Technologies (China)","ror":"https://ror.org/00cmhce21","country_code":"CN","type":"company","lineage":["https://openalex.org/I2250955327"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Fan Yu","raw_affiliation_strings":["Huawei Technologies Co. Ltd.,Hangzhou,China,310000"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Huawei Technologies Co. Ltd.,Hangzhou,China,310000","institution_ids":["https://openalex.org/I2250955327"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5114151973","display_name":"Hongen Shao","orcid":"https://orcid.org/0000-0002-3464-2648"},"institutions":[{"id":"https://openalex.org/I90610280","display_name":"South China University of Technology","ror":"https://ror.org/0530pts50","country_code":"CN","type":"education","lineage":["https://openalex.org/I90610280"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Hongen Shao","raw_affiliation_strings":["School of Future Technology, South China University of Technology,Guangzhou,China,511442"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"School of Future Technology, South China University of Technology,Guangzhou,China,511442","institution_ids":["https://openalex.org/I90610280"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5071736086","display_name":"Xiaofeng Zou","orcid":"https://orcid.org/0000-0002-5823-6345"},"institutions":[{"id":"https://openalex.org/I90610280","display_name":"South China University of Technology","ror":"https://ror.org/0530pts50","country_code":"CN","type":"education","lineage":["https://openalex.org/I90610280"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Xiaofeng Zou","raw_affiliation_strings":["School of Future Technology, South China University of Technology,Guangzhou,China,511442"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"School of Future Technology, South China University of Technology,Guangzhou,China,511442","institution_ids":["https://openalex.org/I90610280"]}]}],"institutions":[],"countries_distinct_count":2,"institutions_distinct_count":4,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":6.3497,"has_fulltext":false,"cited_by_count":5,"citation_normalized_percentile":{"value":0.95994853,"is_in_top_1_percent":false,"is_in_top_10_percent":true},"cited_by_percentile_year":{"min":96,"max":98},"biblio":{"volume":"8","issue":"4","first_page":"779","last_page":"793"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10181","display_name":"Natural Language Processing Techniques","score":0.9872999787330627,"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/T10181","display_name":"Natural Language Processing Techniques","score":0.9872999787330627,"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/T10028","display_name":"Topic Modeling","score":0.9854999780654907,"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/T10201","display_name":"Speech Recognition and Synthesis","score":0.9459999799728394,"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/residual","display_name":"Residual","score":0.7401943206787109},{"id":"https://openalex.org/keywords/decoding-methods","display_name":"Decoding methods","score":0.7046432495117188},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.702434778213501},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.571358859539032},{"id":"https://openalex.org/keywords/language-model","display_name":"Language model","score":0.412275493144989},{"id":"https://openalex.org/keywords/natural-language-processing","display_name":"Natural language processing","score":0.358278751373291},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.3465504050254822},{"id":"https://openalex.org/keywords/algorithm","display_name":"Algorithm","score":0.23862957954406738}],"concepts":[{"id":"https://openalex.org/C155512373","wikidata":"https://www.wikidata.org/wiki/Q287450","display_name":"Residual","level":2,"score":0.7401943206787109},{"id":"https://openalex.org/C57273362","wikidata":"https://www.wikidata.org/wiki/Q576722","display_name":"Decoding methods","level":2,"score":0.7046432495117188},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.702434778213501},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.571358859539032},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.412275493144989},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.358278751373291},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.3465504050254822},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.23862957954406738}],"mesh":[],"locations_count":2,"locations":[{"id":"doi:10.26599/bdma.2024.9020074","is_oa":true,"landing_page_url":"https://doi.org/10.26599/bdma.2024.9020074","pdf_url":null,"source":{"id":"https://openalex.org/S4210209060","display_name":"Big Data Mining and Analytics","issn_l":"2096-0654","issn":["2096-0654","2097-406X"],"is_oa":true,"is_in_doaj":true,"is_core":true,"host_organization":"https://openalex.org/P4310311901","host_organization_name":"Tsinghua University Press","host_organization_lineage":["https://openalex.org/P4310311901"],"host_organization_lineage_names":["Tsinghua University Press"],"type":"journal"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Big Data Mining and Analytics","raw_type":"journal-article"},{"id":"pmh:oai:doaj.org/article:9b7c6bd2e5064f4cbc0f542e5681b461","is_oa":true,"landing_page_url":"https://doaj.org/article/9b7c6bd2e5064f4cbc0f542e5681b461","pdf_url":null,"source":{"id":"https://openalex.org/S4306401280","display_name":"DOAJ (DOAJ: Directory of Open Access Journals)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by-sa","license_id":"https://openalex.org/licenses/cc-by-sa","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"Big Data Mining and Analytics, Vol 8, Iss 4, Pp 779-793 (2025)","raw_type":"article"}],"best_oa_location":{"id":"doi:10.26599/bdma.2024.9020074","is_oa":true,"landing_page_url":"https://doi.org/10.26599/bdma.2024.9020074","pdf_url":null,"source":{"id":"https://openalex.org/S4210209060","display_name":"Big Data Mining and Analytics","issn_l":"2096-0654","issn":["2096-0654","2097-406X"],"is_oa":true,"is_in_doaj":true,"is_core":true,"host_organization":"https://openalex.org/P4310311901","host_organization_name":"Tsinghua University Press","host_organization_lineage":["https://openalex.org/P4310311901"],"host_organization_lineage_names":["Tsinghua University Press"],"type":"journal"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Big Data Mining and Analytics","raw_type":"journal-article"},"sustainable_development_goals":[{"display_name":"Quality Education","id":"https://metadata.un.org/sdg/4","score":0.4300000071525574}],"awards":[{"id":"https://openalex.org/G1125435335","display_name":null,"funder_award_id":"2024ZYGXZR074,2023ZYGXZR023","funder_id":"https://openalex.org/F4320335787","funder_display_name":"Fundamental Research Funds for the Central Universities"},{"id":"https://openalex.org/G2679595405","display_name":null,"funder_award_id":"62406114,62306117","funder_id":"https://openalex.org/F4320321001","funder_display_name":"National Natural Science Foundation of China"},{"id":"https://openalex.org/G6792141703","display_name":null,"funder_award_id":"2021ZD0110400","funder_id":"https://openalex.org/F4320335777","funder_display_name":"National Key Research and Development Program of China"}],"funders":[{"id":"https://openalex.org/F4320321001","display_name":"National Natural Science Foundation of China","ror":"https://ror.org/01h0zpd94"},{"id":"https://openalex.org/F4320335777","display_name":"National Key Research and Development Program of China","ror":null},{"id":"https://openalex.org/F4320335787","display_name":"Fundamental Research Funds for the Central Universities","ror":null}],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":32,"referenced_works":["https://openalex.org/W2783772867","https://openalex.org/W2948947170","https://openalex.org/W3118485687","https://openalex.org/W3136450870","https://openalex.org/W3168126734","https://openalex.org/W4214883719","https://openalex.org/W4247370645","https://openalex.org/W4280641199","https://openalex.org/W4283581972","https://openalex.org/W4385258403","https://openalex.org/W4385571586","https://openalex.org/W4385572299","https://openalex.org/W4389519819","https://openalex.org/W4391582407","https://openalex.org/W4392543779","https://openalex.org/W4395112660","https://openalex.org/W4396506525","https://openalex.org/W4401807361","https://openalex.org/W4402669974","https://openalex.org/W4402672007","https://openalex.org/W4402683901","https://openalex.org/W4402901320","https://openalex.org/W4404250004","https://openalex.org/W6729972426","https://openalex.org/W6754244489","https://openalex.org/W6847386241","https://openalex.org/W6849765749","https://openalex.org/W6857438674","https://openalex.org/W6857696566","https://openalex.org/W6857817228","https://openalex.org/W6860434696","https://openalex.org/W6861670789"],"related_works":["https://openalex.org/W2560215812","https://openalex.org/W2949601986","https://openalex.org/W2788972299","https://openalex.org/W2521347458","https://openalex.org/W2161474341","https://openalex.org/W2498789492","https://openalex.org/W4302615923","https://openalex.org/W3203142394","https://openalex.org/W2351061015","https://openalex.org/W4220731478"],"abstract_inverted_index":{"Large":[0],"language":[1],"Models":[2],"(LLMs)":[3],"have":[4,56],"immense":[5],"potential":[6],"to":[7,18,60,118,148],"enhance":[8],"the":[9,29,39,73,82,108,121,145],"capabilities":[10],"of":[11,33,64,75,81,88,142],"Cyber-Physical-Social":[12],"Intelligence":[13],"(CPSI)":[14],"systems,":[15],"enabling":[16],"them":[17],"better":[19],"engage":[20],"with":[21],"complex":[22],"cyber,":[23],"physical,":[24],"and":[25,103],"social":[26],"environments.":[27],"However,":[28],"high":[30],"inference":[31,70],"latency":[32],"LLMs,":[34],"which":[35],"is":[36],"inherited":[37],"from":[38],"autoregressive":[40,83,150],"decoding":[41,59,77,84,97,105,116],"process,":[42],"hinders":[43],"their":[44],"wide":[45],"application":[46],"in":[47],"CPSI":[48],"systems.":[49],"To":[50],"address":[51],"this":[52],"challenge,":[53],"current":[54],"approaches":[55],"incorporated":[57],"speculative":[58,96],"enable":[61],"parallel":[62],"prediction":[63],"multiple":[65],"subsequent":[66],"tokens,":[67],"thereby":[68],"achieving":[69,137],"acceleration.":[71],"Nevertheless,":[72],"accuracy":[74],"these":[76,89],"heads":[78],"falls":[79],"short":[80],"approach.":[85],"In":[86],"light":[87],"limitations,":[90],"we":[91,112],"propose":[92,113],"ResDecode,":[93],"a":[94,114,138],"novel":[95],"method":[98],"characterized":[99],"by":[100],"its":[101],"efficient":[102],"accurate":[104],"heads.":[106],"Within":[107],"lightweight":[109],"draft":[110],"model,":[111],"residual":[115],"head":[117],"compensate":[119],"for":[120],"full":[122],"context":[123],"encoder's":[124],"limited":[125],"capability":[126],"on":[127,144],"long-range":[128],"dependencies,":[129],"thus":[130],"improving":[131],"accuracy.":[132],"ResDecode":[133],"demonstrates":[134],"impressive":[135],"results,":[136],"maximum":[139],"speedup":[140],"ratio":[141],"3.2\u00d7":[143],"MT-bench":[146],"compared":[147],"vanilla":[149],"decoding.":[151]},"counts_by_year":[{"year":2026,"cited_by_count":2},{"year":2025,"cited_by_count":3}],"updated_date":"2026-07-22T07:51:19.307946","created_date":"2025-10-10T00:00:00"}