{"id":"https://openalex.org/W7134981367","doi":"https://doi.org/10.48550/arxiv.2603.09527","title":"Efficiently Aligning Draft Models via Parameter- and Data-Efficient Adaptation","display_name":"Efficiently Aligning Draft Models via Parameter- and Data-Efficient Adaptation","publication_year":2026,"publication_date":"2026-03-10","ids":{"openalex":"https://openalex.org/W7134981367","doi":"https://doi.org/10.48550/arxiv.2603.09527"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2603.09527","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2603.09527","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"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":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"article"},"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.48550/arxiv.2603.09527","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5100591723","display_name":"Luxi Lin","orcid":null},"institutions":[],"countries":[],"is_corresponding":true,"raw_author_name":"Lin, Luxi","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5128705085","display_name":"Zhihang Lin","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Lin, Zhihang","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5128784734","display_name":"Zhanpeng Zeng","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zeng, Zhanpeng","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5128797418","display_name":"Yuhao Chen","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Chen, Yuhao","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5128718394","display_name":"Qingyu Zhang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhang, Qingyu","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5128726234","display_name":"Jixiang Luo","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Luo, Jixiang","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5128748797","display_name":"Xuelong Li","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Li, Xuelong","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5128757886","display_name":"Rongrong Ji","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ji, Rongrong","raw_affiliation_strings":[],"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":8,"corresponding_author_ids":["https://openalex.org/A5100591723"],"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/T12535","display_name":"Machine Learning and Data Classification","score":0.2842999994754791,"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/T12535","display_name":"Machine Learning and Data Classification","score":0.2842999994754791,"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/T10775","display_name":"Generative Adversarial Networks and Image Synthesis","score":0.16300000250339508,"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/T10028","display_name":"Topic Modeling","score":0.09560000151395798,"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/adaptation","display_name":"Adaptation (eye)","score":0.5960000157356262},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.48980000615119934},{"id":"https://openalex.org/keywords/selection","display_name":"Selection (genetic algorithm)","score":0.4860000014305115},{"id":"https://openalex.org/keywords/sample","display_name":"Sample (material)","score":0.4316999912261963},{"id":"https://openalex.org/keywords/decoding-methods","display_name":"Decoding methods","score":0.39079999923706055},{"id":"https://openalex.org/keywords/throughput","display_name":"Throughput","score":0.37369999289512634},{"id":"https://openalex.org/keywords/code","display_name":"Code (set theory)","score":0.3709000051021576},{"id":"https://openalex.org/keywords/model-selection","display_name":"Model selection","score":0.3659000098705292},{"id":"https://openalex.org/keywords/hull","display_name":"Hull","score":0.36469998955726624}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.646399974822998},{"id":"https://openalex.org/C139807058","wikidata":"https://www.wikidata.org/wiki/Q352374","display_name":"Adaptation (eye)","level":2,"score":0.5960000157356262},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.48980000615119934},{"id":"https://openalex.org/C81917197","wikidata":"https://www.wikidata.org/wiki/Q628760","display_name":"Selection (genetic algorithm)","level":2,"score":0.4860000014305115},{"id":"https://openalex.org/C198531522","wikidata":"https://www.wikidata.org/wiki/Q485146","display_name":"Sample (material)","level":2,"score":0.4316999912261963},{"id":"https://openalex.org/C57273362","wikidata":"https://www.wikidata.org/wiki/Q576722","display_name":"Decoding methods","level":2,"score":0.39079999923706055},{"id":"https://openalex.org/C157764524","wikidata":"https://www.wikidata.org/wiki/Q1383412","display_name":"Throughput","level":3,"score":0.37369999289512634},{"id":"https://openalex.org/C2776760102","wikidata":"https://www.wikidata.org/wiki/Q5139990","display_name":"Code (set theory)","level":3,"score":0.3709000051021576},{"id":"https://openalex.org/C93959086","wikidata":"https://www.wikidata.org/wiki/Q6888345","display_name":"Model selection","level":2,"score":0.3659000098705292},{"id":"https://openalex.org/C37423430","wikidata":"https://www.wikidata.org/wiki/Q6750281","display_name":"Hull","level":2,"score":0.36469998955726624},{"id":"https://openalex.org/C2778915421","wikidata":"https://www.wikidata.org/wiki/Q3643177","display_name":"Performance improvement","level":2,"score":0.35249999165534973},{"id":"https://openalex.org/C123657996","wikidata":"https://www.wikidata.org/wiki/Q12271","display_name":"Architecture","level":2,"score":0.3400999903678894},{"id":"https://openalex.org/C51632099","wikidata":"https://www.wikidata.org/wiki/Q3985153","display_name":"Training set","level":2,"score":0.3165000081062317},{"id":"https://openalex.org/C200601418","wikidata":"https://www.wikidata.org/wiki/Q2193887","display_name":"Reliability engineering","level":1,"score":0.3100999891757965},{"id":"https://openalex.org/C67186912","wikidata":"https://www.wikidata.org/wiki/Q367664","display_name":"Data modeling","level":2,"score":0.30559998750686646},{"id":"https://openalex.org/C135510737","wikidata":"https://www.wikidata.org/wiki/Q860554","display_name":"Performance indicator","level":2,"score":0.30469998717308044},{"id":"https://openalex.org/C2776434776","wikidata":"https://www.wikidata.org/wiki/Q19246213","display_name":"Domain adaptation","level":3,"score":0.3034999966621399},{"id":"https://openalex.org/C2779679103","wikidata":"https://www.wikidata.org/wiki/Q5251805","display_name":"Degradation (telecommunications)","level":2,"score":0.2971999943256378},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.2937000095844269},{"id":"https://openalex.org/C175154964","wikidata":"https://www.wikidata.org/wiki/Q380077","display_name":"Task analysis","level":3,"score":0.29159998893737793},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.2912999987602234},{"id":"https://openalex.org/C2777211547","wikidata":"https://www.wikidata.org/wiki/Q17141490","display_name":"Training (meteorology)","level":2,"score":0.2904999852180481},{"id":"https://openalex.org/C89611455","wikidata":"https://www.wikidata.org/wiki/Q6804646","display_name":"Mechanism (biology)","level":2,"score":0.28189998865127563},{"id":"https://openalex.org/C147764199","wikidata":"https://www.wikidata.org/wiki/Q6865248","display_name":"Minification","level":2,"score":0.2685000002384186},{"id":"https://openalex.org/C98045186","wikidata":"https://www.wikidata.org/wiki/Q205663","display_name":"Process (computing)","level":2,"score":0.26339998841285706},{"id":"https://openalex.org/C127413603","wikidata":"https://www.wikidata.org/wiki/Q11023","display_name":"Engineering","level":0,"score":0.25600001215934753},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.25380000472068787}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2603.09527","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2603.09527","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"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":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"article"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2603.09527","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2603.09527","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"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":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"article"},"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":{"Speculative":[0],"decoding":[1],"accelerates":[2],"LLM":[3],"inference":[4],"but":[5],"suffers":[6],"from":[7],"performance":[8,138],"degradation":[9],"when":[10],"target":[11,28,98],"models":[12,25],"are":[13],"fine-tuned":[14,97,140],"for":[15,26,52,127],"specific":[16],"domains.":[17],"A":[18],"naive":[19],"solution":[20],"is":[21,31,157],"to":[22,71,100,114,153],"retrain":[23],"draft":[24,55],"every":[27],"model,":[29],"which":[30],"costly":[32],"and":[33,42,68,75,110],"inefficient.":[34],"To":[35],"address":[36],"this,":[37],"we":[38],"introduce":[39],"a":[40,62,90,119],"parameter-":[41],"data-efficient":[43],"framework":[44],"named":[45],"Efficient":[46],"Draft":[47],"Adaptation,":[48],"abbreviated":[49],"as":[50],"EDA,":[51],"efficiently":[53],"adapting":[54],"models.":[56],"EDA":[57,134],"introduces":[58],"three":[59],"innovations:":[60],"(1)":[61],"decoupled":[63],"architecture":[64],"that":[65,94,123,133],"utilizes":[66,95],"shared":[67,74],"private":[69,88],"components":[70],"model":[72,99],"the":[73,86,96,106],"target-specific":[76],"output":[77],"distributions":[78],"separately,":[79],"enabling":[80],"parameter-efficient":[81],"adaptation":[82],"by":[83],"updating":[84],"only":[85],"lightweight":[87],"component;(2)":[89],"data":[91,126],"regeneration":[92],"strategy":[93],"regenerate":[101],"training":[102,109,150],"data,":[103],"thereby":[104],"improving":[105],"alignment":[107],"between":[108],"speculative":[111,137],"decoding,":[112],"leading":[113],"higher":[115],"average":[116,144],"acceptance":[117,145],"length;(3)":[118],"sample":[120],"selection":[121],"mechanism":[122],"prioritizes":[124],"high-value":[125],"efficient":[128],"adaptation.":[129],"Our":[130],"experiments":[131],"show":[132],"effectively":[135],"restores":[136],"on":[139],"models,":[141],"achieving":[142],"superior":[143],"lengths":[146],"with":[147],"significantly":[148],"reduced":[149],"costs":[151],"compared":[152],"full":[154],"retraining.":[155],"Code":[156],"available":[158],"at":[159],"https://github.com/Lyn-Lucy/Efficient-Draft-Adaptation.":[160]},"counts_by_year":[],"updated_date":"2026-03-12T06:18:43.230356","created_date":"2026-03-12T00:00:00"}