{"id":"https://openalex.org/W7166853010","doi":"https://doi.org/10.18653/v1/2026.findings-acl.802","title":"SpecBound: Adaptive Bounded Self-Speculation with Layer-wise Confidence Calibration","display_name":"SpecBound: Adaptive Bounded Self-Speculation with Layer-wise Confidence Calibration","publication_year":2026,"publication_date":"2026-01-01","ids":{"openalex":"https://openalex.org/W7166853010","doi":"https://doi.org/10.18653/v1/2026.findings-acl.802"},"language":null,"primary_location":{"id":"doi:10.18653/v1/2026.findings-acl.802","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.findings-acl.802","pdf_url":"https://aclanthology.org/2026.findings-acl.802.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":"Findings of the Association for Computational Linguistics: ACL 2026","raw_type":"proceedings-article"},"type":"conference-paper","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://aclanthology.org/2026.findings-acl.802.pdf","any_repository_has_fulltext":null},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5139769591","display_name":"Zhuofan Wen","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhuofan Wen","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5139808511","display_name":"Yang Feng","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yang Feng","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":0.0,"has_fulltext":true,"cited_by_count":0,"citation_normalized_percentile":{"value":0.82121896,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"16309","last_page":"16321"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10054","display_name":"Parallel Computing and Optimization Techniques","score":0.15070000290870667,"subfield":{"id":"https://openalex.org/subfields/1708","display_name":"Hardware and Architecture"},"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/T10054","display_name":"Parallel Computing and Optimization Techniques","score":0.15070000290870667,"subfield":{"id":"https://openalex.org/subfields/1708","display_name":"Hardware and Architecture"},"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/T10901","display_name":"Advanced Data Compression Techniques","score":0.07360000163316727,"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/T10904","display_name":"Embedded Systems Design Techniques","score":0.053300000727176666,"subfield":{"id":"https://openalex.org/subfields/1708","display_name":"Hardware and Architecture"},"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/calibration","display_name":"Calibration","score":0.60589998960495},{"id":"https://openalex.org/keywords/bounded-function","display_name":"Bounded function","score":0.5633999705314636},{"id":"https://openalex.org/keywords/noise","display_name":"Noise (video)","score":0.3082999885082245},{"id":"https://openalex.org/keywords/control-theory","display_name":"Control theory (sociology)","score":0.27889999747276306},{"id":"https://openalex.org/keywords/interval","display_name":"Interval (graph theory)","score":0.27300000190734863}],"concepts":[{"id":"https://openalex.org/C165838908","wikidata":"https://www.wikidata.org/wiki/Q736777","display_name":"Calibration","level":2,"score":0.60589998960495},{"id":"https://openalex.org/C34388435","wikidata":"https://www.wikidata.org/wiki/Q2267362","display_name":"Bounded function","level":2,"score":0.5633999705314636},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.45260000228881836},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.4453999996185303},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.43309998512268066},{"id":"https://openalex.org/C99498987","wikidata":"https://www.wikidata.org/wiki/Q2210247","display_name":"Noise (video)","level":3,"score":0.3082999885082245},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.299699991941452},{"id":"https://openalex.org/C47446073","wikidata":"https://www.wikidata.org/wiki/Q5165890","display_name":"Control theory (sociology)","level":3,"score":0.27889999747276306},{"id":"https://openalex.org/C2778067643","wikidata":"https://www.wikidata.org/wiki/Q166507","display_name":"Interval (graph theory)","level":2,"score":0.27300000190734863},{"id":"https://openalex.org/C126255220","wikidata":"https://www.wikidata.org/wiki/Q141495","display_name":"Mathematical optimization","level":1,"score":0.27160000801086426},{"id":"https://openalex.org/C143380155","wikidata":"https://www.wikidata.org/wiki/Q5160088","display_name":"Confidence region","level":3,"score":0.25999999046325684},{"id":"https://openalex.org/C61797465","wikidata":"https://www.wikidata.org/wiki/Q1188986","display_name":"Term (time)","level":2,"score":0.2500999867916107}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.18653/v1/2026.findings-acl.802","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.findings-acl.802","pdf_url":"https://aclanthology.org/2026.findings-acl.802.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":"Findings of the Association for Computational Linguistics: ACL 2026","raw_type":"proceedings-article"}],"best_oa_location":{"id":"doi:10.18653/v1/2026.findings-acl.802","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.findings-acl.802","pdf_url":"https://aclanthology.org/2026.findings-acl.802.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":"Findings of the Association for Computational Linguistics: ACL 2026","raw_type":"proceedings-article"},"sustainable_development_goals":[{"id":"https://metadata.un.org/sdg/10","display_name":"Reduced inequalities","score":0.4241086542606354}],"awards":[],"funders":[],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7166853010.pdf","grobid_xml":"https://content.openalex.org/works/W7166853010.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Speculative":[0],"decoding":[1,96,145],"has":[2],"emerged":[3],"as":[4],"a":[5,52,73,106],"promising":[6],"approach":[7],"to":[8,130,138],"accelerate":[9],"autoregressive":[10,144],"inference":[11],"in":[12,51,85,105],"large":[13],"language":[14],"models":[15,32],"(LLMs).Selfdraft":[16],"methods,":[17],"which":[18],"leverage":[19],"the":[20,27,46,99,120,131],"base":[21,132],"LLM":[22,133],"itself":[23],"for":[24],"speculation,":[25],"avoid":[26],"overhead":[28],"of":[29,48,102],"auxiliary":[30],"draft":[31,53,63,103],"but":[33],"face":[34],"limitations:":[35],"shallow":[36],"layers":[37],"often":[38],"produce":[39],"overconfident":[40],"yet":[41],"incorrect":[42],"token":[43],"predictions,":[44],"and":[45,65,88,135,151],"presence":[47],"difficult":[49],"tokens":[50,104],"sequence":[54],"forces":[55],"redundant":[56],"computation":[57],"through":[58,110],"deeper":[59],"layers,":[60,112],"undermining":[61],"both":[62],"acceptance":[64],"overall":[66],"speedup.To":[67],"address":[68],"these":[69],"issues,":[70],"we":[71],"propose":[72],"novel":[74],"self-draft":[75],"framework":[76],"that":[77],"suppresses":[78],"spurious":[79],"confidence":[80],"via":[81],"layer-wise":[82],"temperature":[83],"annealing":[84],"early-exit":[86],"decision":[87],"adaptively":[89],"bounds":[90],"speculation":[91],"length":[92],"based":[93],"on":[94],"token-wise":[95],"difficulty.By":[97],"reprocessing":[98],"hidden":[100],"states":[101],"unified":[107],"parallel":[108],"pass":[109],"deep":[111],"our":[113],"method":[114],"maintains":[115],"exact":[116],"output":[117],"equivalence":[118],"with":[119],"original":[121],"model":[122,153],"while":[123],"maximizing":[124],"computational":[125],"efficiency.It":[126],"requires":[127],"no":[128],"modifications":[129],"parameters":[134],"achieves":[136],"up":[137],"2.33\u00d7":[139],"wall-time":[140],"speedup":[141],"over":[142],"standard":[143],"across":[146],"diverse":[147],"long-form":[148],"generation":[149],"tasks":[150],"multiple":[152],"architectures.":[154]},"counts_by_year":[],"updated_date":"2026-07-29T14:22:42.915294","created_date":"2026-07-02T00:00:00"}