{"id":"https://openalex.org/W7159600725","doi":"https://doi.org/10.48550/arxiv.2604.27747","title":"Position-Aware Drafting for Inference Acceleration in LLM-Based Generative List-Wise Recommendation","display_name":"Position-Aware Drafting for Inference Acceleration in LLM-Based Generative List-Wise Recommendation","publication_year":2026,"publication_date":"2026-04-30","ids":{"openalex":"https://openalex.org/W7159600725","doi":"https://doi.org/10.48550/arxiv.2604.27747"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2604.27747","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.27747","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"},"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.48550/arxiv.2604.27747","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5061871626","display_name":"Jiaju Chen","orcid":"https://orcid.org/0009-0000-1940-1515"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Chen, Jiaju","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5134977852","display_name":"Chongming Gao","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Gao, Chongming","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5084334472","display_name":"Chenxiao Fan","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Fan, Chenxiao","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5134945210","display_name":"Haoyan Liu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Liu, Haoyan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5134983650","display_name":"Qingpeng Cai","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Cai, Qingpeng","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5134964799","display_name":"Peng Jiang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Jiang, Peng","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5134974479","display_name":"Xiangnan He","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"He, Xiangnan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":0,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":null,"last_page":null},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10203","display_name":"Recommender Systems and Techniques","score":0.3522000014781952,"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":0.3522000014781952,"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/T10775","display_name":"Generative Adversarial Networks and Image Synthesis","score":0.10909999907016754,"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/T13702","display_name":"Machine Learning in Healthcare","score":0.07729999721050262,"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/inference","display_name":"Inference","score":0.7594000101089478},{"id":"https://openalex.org/keywords/speedup","display_name":"Speedup","score":0.7376999855041504},{"id":"https://openalex.org/keywords/generative-grammar","display_name":"Generative grammar","score":0.597000002861023},{"id":"https://openalex.org/keywords/semantics","display_name":"Semantics (computer science)","score":0.546500027179718},{"id":"https://openalex.org/keywords/decoding-methods","display_name":"Decoding methods","score":0.542900025844574},{"id":"https://openalex.org/keywords/generative-model","display_name":"Generative model","score":0.5353999733924866},{"id":"https://openalex.org/keywords/encode","display_name":"ENCODE","score":0.48500001430511475},{"id":"https://openalex.org/keywords/key","display_name":"Key (lock)","score":0.43560001254081726}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7858999967575073},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.7594000101089478},{"id":"https://openalex.org/C68339613","wikidata":"https://www.wikidata.org/wiki/Q1549489","display_name":"Speedup","level":2,"score":0.7376999855041504},{"id":"https://openalex.org/C39890363","wikidata":"https://www.wikidata.org/wiki/Q36108","display_name":"Generative grammar","level":2,"score":0.597000002861023},{"id":"https://openalex.org/C184337299","wikidata":"https://www.wikidata.org/wiki/Q1437428","display_name":"Semantics (computer science)","level":2,"score":0.546500027179718},{"id":"https://openalex.org/C57273362","wikidata":"https://www.wikidata.org/wiki/Q576722","display_name":"Decoding methods","level":2,"score":0.542900025844574},{"id":"https://openalex.org/C167966045","wikidata":"https://www.wikidata.org/wiki/Q5532625","display_name":"Generative model","level":3,"score":0.5353999733924866},{"id":"https://openalex.org/C66746571","wikidata":"https://www.wikidata.org/wiki/Q1134833","display_name":"ENCODE","level":3,"score":0.48500001430511475},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.43560001254081726},{"id":"https://openalex.org/C198082294","wikidata":"https://www.wikidata.org/wiki/Q3399648","display_name":"Position (finance)","level":2,"score":0.396699994802475},{"id":"https://openalex.org/C186644900","wikidata":"https://www.wikidata.org/wiki/Q194152","display_name":"Parsing","level":2,"score":0.3939000070095062},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.39149999618530273},{"id":"https://openalex.org/C2780451532","wikidata":"https://www.wikidata.org/wiki/Q759676","display_name":"Task (project management)","level":2,"score":0.3774000108242035},{"id":"https://openalex.org/C48145219","wikidata":"https://www.wikidata.org/wiki/Q1335365","display_name":"Security token","level":2,"score":0.37209999561309814},{"id":"https://openalex.org/C125411270","wikidata":"https://www.wikidata.org/wiki/Q18653","display_name":"Encoding (memory)","level":2,"score":0.35339999198913574},{"id":"https://openalex.org/C2780586882","wikidata":"https://www.wikidata.org/wiki/Q7520643","display_name":"Simple (philosophy)","level":2,"score":0.34389999508857727},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.3206000030040741},{"id":"https://openalex.org/C42058472","wikidata":"https://www.wikidata.org/wiki/Q810214","display_name":"Base (topology)","level":2,"score":0.3100000023841858},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.3091999888420105},{"id":"https://openalex.org/C117896860","wikidata":"https://www.wikidata.org/wiki/Q11376","display_name":"Acceleration","level":2,"score":0.2978000044822693},{"id":"https://openalex.org/C80444323","wikidata":"https://www.wikidata.org/wiki/Q2878974","display_name":"Theoretical computer science","level":1,"score":0.2971000075340271},{"id":"https://openalex.org/C195324797","wikidata":"https://www.wikidata.org/wiki/Q33742","display_name":"Natural language","level":2,"score":0.28859999775886536},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.2694999873638153},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.26649999618530273},{"id":"https://openalex.org/C146499914","wikidata":"https://www.wikidata.org/wiki/Q5469969","display_name":"Formal semantics (linguistics)","level":2,"score":0.26249998807907104},{"id":"https://openalex.org/C199360897","wikidata":"https://www.wikidata.org/wiki/Q9143","display_name":"Programming language","level":1,"score":0.25049999356269836}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2604.27747","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.27747","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2604.27747","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.27747","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Large":[0],"language":[1],"model":[2,32,126,154],"(LLM)-based":[3],"generative":[4,57,117],"list-wise":[5],"recommendation":[6,230],"has":[7],"advanced":[8],"rapidly,":[9],"but":[10],"decoding":[11,26],"remains":[12],"sequential":[13],"and":[14,40,46,71,93,160,181,199,216],"thus":[15],"latency-prone.":[16],"To":[17,164],"accelerate":[18],"inference":[19,202],"without":[20],"changing":[21],"the":[22,48,124,136,149,153],"target":[23,42],"distribution,":[24],"speculative":[25],"(SD)":[27],"uses":[28],"a":[29,41,85,119,175,182],"small":[30],"draft":[31,125,150,186,197],"to":[33,44,97,155,157,193,212],"propose":[34],"several":[35],"next":[36],"tokens":[37,77],"at":[38],"once":[39],"LLM":[43],"verify":[45],"accept":[47],"longest":[49],"prefix,":[50],"skipping":[51],"multiple":[52,65],"steps":[53],"per":[54],"round.":[55],"In":[56],"recommendation,":[58],"however,":[59],"each":[60,140],"item":[61,179],"is":[62,190],"represented":[63],"by":[64],"semantic-ID":[66],"tokens,":[67],"often":[68],"with":[69,99,127,168,195],"separators,":[70],"current":[72],"drafts":[73],"typically":[74],"treat":[75],"these":[76,104,166],"uniformly.":[78],"This":[79],"overlooks":[80],"two":[81,128],"practical":[82],"facts:":[83],"(i)":[84],"token's":[86],"semantics":[87],"depend":[88],"on":[89,206],"its":[90],"within-item":[91,137],"slot,":[92],"(ii)":[94],"uncertainty":[95,159],"tends":[96],"increase":[98],"speculation":[100],"depth.":[101],"Without":[102],"modeling":[103],"effects,":[105],"SD's":[106],"speedups":[107],"can":[108],"be":[109],"limited.":[110],"We":[111],"introduce":[112],"PAD-Rec,":[113],"Position-Aware":[114],"Drafting":[115],"for":[116,178,185],"Recommendation,":[118],"lightweight":[120],"module":[121,189],"that":[122],"augments":[123],"complementary":[129],"signals.":[130],"Item":[131],"position":[132,146],"embeddings":[133,147],"explicitly":[134],"encode":[135,148],"slot":[138],"of":[139],"token,":[141],"strengthening":[142],"structural":[143],"awareness.":[144],"Step":[145],"step,":[151],"allowing":[152],"adapt":[156],"depth-dependent":[158],"improve":[161],"proposal":[162],"quality.":[163,231],"harmonize":[165],"signals":[167],"base":[169],"features,":[170],"we":[171],"add":[172],"simple":[173],"gates:":[174],"learnable":[176],"coefficient":[177],"slots":[180],"context-driven":[183],"gate":[184],"steps.":[187],"The":[188],"trainable,":[191],"easy":[192],"integrate":[194],"standard":[196],"models,":[198],"adds":[200],"negligible":[201],"overhead.":[203],"Extensive":[204],"experiments":[205],"four":[207],"real-world":[208],"datasets":[209],"show":[210],"up":[211],"3.1x":[213],"wall-clock":[214,220],"speedup":[215,221],"about":[217],"5%":[218],"average":[219],"gain":[222],"over":[223],"strong":[224],"SD":[225],"baselines,":[226],"while":[227],"largely":[228],"preserving":[229]},"counts_by_year":[],"updated_date":"2026-07-28T07:46:37.118299","created_date":"2026-05-02T00:00:00"}