{"id":"https://openalex.org/W4412394824","doi":"https://doi.org/10.1145/3726302.3729961","title":"Efficiency Unleashed: Inference Acceleration for LLM-based Recommender Systems with Speculative Decoding","display_name":"Efficiency Unleashed: Inference Acceleration for LLM-based Recommender Systems with Speculative Decoding","publication_year":2025,"publication_date":"2025-07-13","ids":{"openalex":"https://openalex.org/W4412394824","doi":"https://doi.org/10.1145/3726302.3729961"},"language":"en","primary_location":{"id":"doi:10.1145/3726302.3729961","is_oa":false,"landing_page_url":"https://doi.org/10.1145/3726302.3729961","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 48th International ACM SIGIR Conference on Research and Development in Information Retrieval","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/A5013513956","display_name":"Yunjia Xi","orcid":"https://orcid.org/0000-0001-6883-881X"},"institutions":[{"id":"https://openalex.org/I183067930","display_name":"Shanghai Jiao Tong University","ror":"https://ror.org/0220qvk04","country_code":"CN","type":"education","lineage":["https://openalex.org/I183067930"]}],"countries":["CN"],"is_corresponding":true,"raw_author_name":"Yunjia Xi","raw_affiliation_strings":["Shanghai Jiao Tong University, Shanghai, China"],"affiliations":[{"raw_affiliation_string":"Shanghai Jiao Tong University, Shanghai, China","institution_ids":["https://openalex.org/I183067930"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5032497440","display_name":"Hangyu Wang","orcid":"https://orcid.org/0009-0006-0410-5426"},"institutions":[{"id":"https://openalex.org/I183067930","display_name":"Shanghai Jiao Tong University","ror":"https://ror.org/0220qvk04","country_code":"CN","type":"education","lineage":["https://openalex.org/I183067930"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Hangyu Wang","raw_affiliation_strings":["Shanghai Jiao Tong University, Shanghai, China"],"affiliations":[{"raw_affiliation_string":"Shanghai Jiao Tong University, Shanghai, China","institution_ids":["https://openalex.org/I183067930"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100427434","display_name":"Bo Chen","orcid":"https://orcid.org/0000-0003-3750-2533"},"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":"Bo Chen","raw_affiliation_strings":["Huawei Noah's Ark Lab, Shanghai, China"],"affiliations":[{"raw_affiliation_string":"Huawei Noah's Ark Lab, Shanghai, China","institution_ids":["https://openalex.org/I2250955327"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5036057873","display_name":"Jianghao Lin","orcid":"https://orcid.org/0000-0002-8953-3203"},"institutions":[{"id":"https://openalex.org/I183067930","display_name":"Shanghai Jiao Tong University","ror":"https://ror.org/0220qvk04","country_code":"CN","type":"education","lineage":["https://openalex.org/I183067930"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Jianghao Lin","raw_affiliation_strings":["Shanghai Jiao Tong University, Shanghai, China"],"affiliations":[{"raw_affiliation_string":"Shanghai Jiao Tong University, Shanghai, China","institution_ids":["https://openalex.org/I183067930"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5034417837","display_name":"Menghui Zhu","orcid":"https://orcid.org/0000-0002-8567-2185"},"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":"Menghui Zhu","raw_affiliation_strings":["Huawei Noah's Ark Lab, Shanghai, China"],"affiliations":[{"raw_affiliation_string":"Huawei Noah's Ark Lab, Shanghai, China","institution_ids":["https://openalex.org/I2250955327"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101500385","display_name":"Weiwen Liu","orcid":"https://orcid.org/0000-0002-9148-3997"},"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":"Weiwen Liu","raw_affiliation_strings":["Huawei Noah's Ark Lab, Shenzhen, China"],"affiliations":[{"raw_affiliation_string":"Huawei Noah's Ark Lab, Shenzhen, China","institution_ids":["https://openalex.org/I2250955327"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5054330014","display_name":"Ruiming Tang","orcid":"https://orcid.org/0000-0002-9224-2431"},"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":"Ruiming Tang","raw_affiliation_strings":["Huawei Noah's Ark Lab, Shenzhen, China"],"affiliations":[{"raw_affiliation_string":"Huawei Noah's Ark Lab, Shenzhen, China","institution_ids":["https://openalex.org/I2250955327"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5074858555","display_name":"Zhewei Wei","orcid":"https://orcid.org/0000-0003-3620-5086"},"institutions":[{"id":"https://openalex.org/I78988378","display_name":"Renmin University of China","ror":"https://ror.org/041pakw92","country_code":"CN","type":"education","lineage":["https://openalex.org/I78988378"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Zhewei Wei","raw_affiliation_strings":["Renmin University of China, Beijing, China"],"affiliations":[{"raw_affiliation_string":"Renmin University of China, Beijing, China","institution_ids":["https://openalex.org/I78988378"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5090720315","display_name":"Weinan Zhang","orcid":"https://orcid.org/0000-0002-0127-2425"},"institutions":[{"id":"https://openalex.org/I183067930","display_name":"Shanghai Jiao Tong University","ror":"https://ror.org/0220qvk04","country_code":"CN","type":"education","lineage":["https://openalex.org/I183067930"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Weinan Zhang","raw_affiliation_strings":["Shanghai Jiao Tong University, Shanghai, China"],"affiliations":[{"raw_affiliation_string":"Shanghai Jiao Tong University, Shanghai, China","institution_ids":["https://openalex.org/I183067930"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5001571390","display_name":"Yong Yu","orcid":"https://orcid.org/0000-0003-0281-8271"},"institutions":[{"id":"https://openalex.org/I183067930","display_name":"Shanghai Jiao Tong University","ror":"https://ror.org/0220qvk04","country_code":"CN","type":"education","lineage":["https://openalex.org/I183067930"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Yong Yu","raw_affiliation_strings":["Shanghai Jiao Tong University, Shanghai, China"],"affiliations":[{"raw_affiliation_string":"Shanghai Jiao Tong University, Shanghai, China","institution_ids":["https://openalex.org/I183067930"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":10,"corresponding_author_ids":["https://openalex.org/A5013513956"],"corresponding_institution_ids":["https://openalex.org/I183067930"],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":{"value":0.25256941,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"1891","last_page":"1901"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10203","display_name":"Recommender Systems and Techniques","score":0.9993000030517578,"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"}},"topics":[{"id":"https://openalex.org/T10203","display_name":"Recommender Systems and Techniques","score":0.9993000030517578,"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"}},{"id":"https://openalex.org/T10824","display_name":"Image Retrieval and Classification Techniques","score":0.9937999844551086,"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"}},{"id":"https://openalex.org/T10627","display_name":"Advanced Image and Video Retrieval Techniques","score":0.9836999773979187,"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/recommender-system","display_name":"Recommender system","score":0.7700895071029663},{"id":"https://openalex.org/keywords/acceleration","display_name":"Acceleration","score":0.7381864786148071},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.7340912222862244},{"id":"https://openalex.org/keywords/decoding-methods","display_name":"Decoding methods","score":0.6787075996398926},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.659202516078949},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.31870192289352417},{"id":"https://openalex.org/keywords/machine-learning","display_name":"Machine learning","score":0.28851407766342163},{"id":"https://openalex.org/keywords/algorithm","display_name":"Algorithm","score":0.20764762163162231}],"concepts":[{"id":"https://openalex.org/C557471498","wikidata":"https://www.wikidata.org/wiki/Q554950","display_name":"Recommender system","level":2,"score":0.7700895071029663},{"id":"https://openalex.org/C117896860","wikidata":"https://www.wikidata.org/wiki/Q11376","display_name":"Acceleration","level":2,"score":0.7381864786148071},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7340912222862244},{"id":"https://openalex.org/C57273362","wikidata":"https://www.wikidata.org/wiki/Q576722","display_name":"Decoding methods","level":2,"score":0.6787075996398926},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.659202516078949},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.31870192289352417},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.28851407766342163},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.20764762163162231},{"id":"https://openalex.org/C121332964","wikidata":"https://www.wikidata.org/wiki/Q413","display_name":"Physics","level":0,"score":0.0},{"id":"https://openalex.org/C74650414","wikidata":"https://www.wikidata.org/wiki/Q11397","display_name":"Classical mechanics","level":1,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1145/3726302.3729961","is_oa":false,"landing_page_url":"https://doi.org/10.1145/3726302.3729961","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 48th International ACM SIGIR Conference on Research and Development in Information Retrieval","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[{"score":0.5,"display_name":"Affordable and clean energy","id":"https://metadata.un.org/sdg/7"}],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":30,"referenced_works":["https://openalex.org/W2016995838","https://openalex.org/W2103868202","https://openalex.org/W2723293840","https://openalex.org/W3045200674","https://openalex.org/W4213069590","https://openalex.org/W4290790999","https://openalex.org/W4327810158","https://openalex.org/W4379382506","https://openalex.org/W4385562639","https://openalex.org/W4386728933","https://openalex.org/W4389519819","https://openalex.org/W4392384703","https://openalex.org/W4392846385","https://openalex.org/W4395112660","https://openalex.org/W4396722619","https://openalex.org/W4396723256","https://openalex.org/W4396734745","https://openalex.org/W4396757491","https://openalex.org/W4396758712","https://openalex.org/W4396758737","https://openalex.org/W4396843600","https://openalex.org/W4400118952","https://openalex.org/W4400524512","https://openalex.org/W4400524801","https://openalex.org/W4400526723","https://openalex.org/W4400529373","https://openalex.org/W4400909953","https://openalex.org/W4403220034","https://openalex.org/W4403220611","https://openalex.org/W4404639422"],"related_works":["https://openalex.org/W4390273403","https://openalex.org/W4386781444","https://openalex.org/W2150182025","https://openalex.org/W3092950680","https://openalex.org/W3197542405","https://openalex.org/W2056712470","https://openalex.org/W3125580266","https://openalex.org/W4288390103","https://openalex.org/W4317039510","https://openalex.org/W4238861846"],"abstract_inverted_index":{"The":[0],"past":[1],"few":[2],"years":[3],"have":[4],"witnessed":[5],"a":[6,20,74,97,154,175,194],"growing":[7],"interest":[8],"in":[9,19,39,120],"LLM-based":[10,148],"recommender":[11],"systems":[12],"(RSs),":[13],"although":[14],"their":[15],"industrial":[16],"deployment":[17],"remains":[18],"preliminary":[21],"stage.":[22],"Most":[23],"existing":[24],"deployments":[25],"leverage":[26],"LLMs":[27,52],"offline":[28,48],"as":[29,96],"feature":[30],"enhancers,":[31],"generating":[32],"augmented":[33],"knowledge":[34,49,94],"for":[35,102,133,147],"downstream":[36,200],"tasks.":[37],"However,":[38],"recommendation":[40,93,201],"scenarios":[41],"with":[42,51,198],"numerous":[43],"users":[44,119],"and":[45,56,118,126,162,181],"items,":[46],"even":[47],"generation":[50,95],"demands":[53],"significant":[54],"time":[55],"computational":[57,186],"resources.":[58],"This":[59],"inefficiency":[60],"arises":[61],"from":[62],"the":[63,79,113,167,189],"autoregressive":[64],"nature":[65],"of":[66,81,116,170,185],"LLMs.":[67],"A":[68],"promising":[69],"solution":[70],"is":[71,205],"speculative":[72,104],"decoding,":[73],"Draft-Then-Verify":[75],"approach":[76],"that":[77],"increases":[78],"number":[80,115],"tokens":[82],"generated":[83],"per":[84],"decoding":[85],"step.":[86],"In":[87],"this":[88],"work,":[89],"we":[90,107,140],"first":[91],"identify":[92],"highly":[98],"fitting":[99],"use":[100],"case":[101],"retrieval-based":[103],"decoding.":[105],"Then,":[106],"discern":[108],"its":[109],"two":[110],"characteristics:":[111],"(1)":[112],"vast":[114],"items":[117],"RSs":[121,128],"leads":[122],"to":[123,158,165],"retrieval":[124,160],"inefficiency,":[125],"(2)":[127],"exhibit":[129],"high":[130],"diversity":[131],"tolerance":[132],"LLM-generated":[134],"text.":[135],"Building":[136],"on":[137,178,193],"these":[138],"insights,":[139],"introduce":[141],"Lossless":[142],"Acceleration":[143],"via":[144],"Speculative":[145],"Decoding":[146],"Recommender":[149],"Systems":[150],"(LASER),":[151],"which":[152],"features":[153],"Customized":[155],"Retrieval":[156],"Pool":[157],"enhance":[159],"efficiency":[161],"Relaxed":[163],"Verification":[164],"improve":[166],"acceptance":[168],"rate":[169],"draft":[171],"tokens.":[172],"LASER":[173],"achieves":[174],"3-5x":[176],"speedup":[177],"public":[179],"datasets":[180],"saves":[182],"about":[183],"67%":[184],"resources":[187],"during":[188],"online":[190],"A/B":[191],"test":[192],"large-scale":[195],"advertising":[196],"scenario":[197],"lossless":[199],"performance.":[202],"Our":[203],"code":[204],"available":[206],"at":[207],"https://github.com/YunjiaXi/LASER":[208]},"counts_by_year":[],"updated_date":"2026-03-27T05:58:40.876381","created_date":"2025-10-10T00:00:00"}