{"id":"https://openalex.org/W7166819624","doi":"https://doi.org/10.18653/v1/2026.acl-long.1046","title":"SADA: Bridging In-Context Learning and Fine-Tuning via State-Aligned Distillation Adapters","display_name":"SADA: Bridging In-Context Learning and Fine-Tuning via State-Aligned Distillation Adapters","publication_year":2026,"publication_date":"2026-01-01","ids":{"openalex":"https://openalex.org/W7166819624","doi":"https://doi.org/10.18653/v1/2026.acl-long.1046"},"language":null,"primary_location":{"id":"doi:10.18653/v1/2026.acl-long.1046","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.acl-long.1046","pdf_url":"https://aclanthology.org/2026.acl-long.1046.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":"Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)","raw_type":"proceedings-article"},"type":"conference-paper","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://aclanthology.org/2026.acl-long.1046.pdf","any_repository_has_fulltext":null},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5139740351","display_name":"Wenhao Gao","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wenhao Gao","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139775378","display_name":"Tianlong Wang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Tianlong Wang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139802621","display_name":"Wei Jia","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wei Jia","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5014957302","display_name":"Linhao Zhang","orcid":"https://orcid.org/0000-0002-8484-1733"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Linhao Zhang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139823058","display_name":"Aiwei Liu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Aiwei Liu","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5075331030","display_name":"Miao Fan","orcid":"https://orcid.org/0000-0002-0377-9860"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Miao Fan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5114218225","display_name":"Zhou Xiao","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhou Xiao","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":0.80928822,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"22847","last_page":"22862"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T11307","display_name":"Domain Adaptation and Few-Shot Learning","score":0.3041999936103821,"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/T11307","display_name":"Domain Adaptation and Few-Shot Learning","score":0.3041999936103821,"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/T12761","display_name":"Data Stream Mining Techniques","score":0.06880000233650208,"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/T10444","display_name":"Context-Aware Activity Recognition Systems","score":0.04859999939799309,"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/bridging","display_name":"Bridging (networking)","score":0.7197999954223633},{"id":"https://openalex.org/keywords/distillation","display_name":"Distillation","score":0.46880000829696655},{"id":"https://openalex.org/keywords/fractionating-column","display_name":"Fractionating column","score":0.28679999709129333},{"id":"https://openalex.org/keywords/mode","display_name":"Mode (computer interface)","score":0.25270000100135803}],"concepts":[{"id":"https://openalex.org/C174348530","wikidata":"https://www.wikidata.org/wiki/Q188635","display_name":"Bridging (networking)","level":2,"score":0.7197999954223633},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.5896000266075134},{"id":"https://openalex.org/C204030448","wikidata":"https://www.wikidata.org/wiki/Q101017","display_name":"Distillation","level":2,"score":0.46880000829696655},{"id":"https://openalex.org/C21880701","wikidata":"https://www.wikidata.org/wiki/Q2144042","display_name":"Process engineering","level":1,"score":0.4041000008583069},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.32010000944137573},{"id":"https://openalex.org/C127413603","wikidata":"https://www.wikidata.org/wiki/Q11023","display_name":"Engineering","level":0,"score":0.28690001368522644},{"id":"https://openalex.org/C154030694","wikidata":"https://www.wikidata.org/wiki/Q1436074","display_name":"Fractionating column","level":3,"score":0.28679999709129333},{"id":"https://openalex.org/C199639397","wikidata":"https://www.wikidata.org/wiki/Q1788588","display_name":"Engineering drawing","level":1,"score":0.2741999924182892},{"id":"https://openalex.org/C48677424","wikidata":"https://www.wikidata.org/wiki/Q6888088","display_name":"Mode (computer interface)","level":2,"score":0.25270000100135803},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.2484000027179718}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.18653/v1/2026.acl-long.1046","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.acl-long.1046","pdf_url":"https://aclanthology.org/2026.acl-long.1046.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":"Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)","raw_type":"proceedings-article"}],"best_oa_location":{"id":"doi:10.18653/v1/2026.acl-long.1046","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.acl-long.1046","pdf_url":"https://aclanthology.org/2026.acl-long.1046.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":"Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)","raw_type":"proceedings-article"},"sustainable_development_goals":[],"awards":[],"funders":[{"id":"https://openalex.org/F4320316083","display_name":"Tencent","ror":"https://ror.org/00hhjss72"}],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7166819624.pdf","grobid_xml":"https://content.openalex.org/works/W7166819624.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Prompt-based":[0],"in-context":[1],"learning":[2],"(ICL)":[3],"and":[4,76,101,112,121,124,136,148,158],"parameter":[5],"fine-tuning":[6],"are":[7],"two":[8,71],"dominant":[9],"paradigms":[10],"for":[11],"incorporating":[12],"external":[13],"information":[14],"into":[15,37],"large":[16],"language":[17,118],"models":[18],"(LLMs),":[19],"but":[20],"they":[21],"incur":[22],"high":[23],"inference":[24],"costs":[25],"or":[26],"require":[27],"expensive":[28],"retraining.To":[29],"bridge":[30],"this":[31,68],"gap,":[32],"context-to-parameter":[33],"mapping":[34],"converts":[35],"prompts":[36],"temporary":[38],"adapter":[39],"weights.However,":[40],"we":[41,82],"identify":[42],"a":[43,93],"critical":[44],"failure":[45,69],"mode":[46],"in":[47,64],"existing":[48],"methods:":[49],"hiddenstate":[50],"collapse,":[51],"where":[52],"the":[53,61,89,109,113],"adapter-augmented":[54,110],"model's":[55],"internal":[56],"states":[57],"diverge":[58],"sharply":[59],"from":[60],"full-context":[62,114],"oracle":[63],"deeper":[65],"layers.We":[66],"trace":[67],"to":[70,97,105,141],"coupled":[72],"gaps:":[73],"suboptimal":[74],"Input-Selection":[75],"inadequate":[77],"Supervision-Signal.To":[78],"address":[79],"these":[80],"issues,":[81],"propose":[83],"SADA":[84,129],"(State-Aligned":[85],"Distillation":[86],"Adapters).We":[87],"establish":[88],"attention-block":[90],"output":[91],"as":[92],"principled":[94],"feature":[95],"interface":[96],"improve":[98],"input":[99],"selection":[100],"introduce":[102],"statealignment":[103],"distillation":[104],"enforce":[106],"consistency":[107],"between":[108],"model":[111],"oracle.Experiments":[115],"on":[116],"long-context":[117],"modeling":[119],"(PG19)":[120],"downstream":[122],"NLU":[123],"summarization":[125],"benchmarks":[126],"show":[127],"that":[128],"consistently":[130],"outperforms":[131],"strong":[132],"baselines":[133],"like":[134],"StreamAdapter":[135],"GenerativeAdapter,":[137],"achieving":[138],"performance":[139],"comparable":[140],"ICL":[142],"while":[143],"significantly":[144],"reducing":[145],"memory":[146],"footprint":[147],"latency.We":[149],"further":[150],"analyze":[151],"when":[152,159],"parameterized":[153],"context":[154,161],"compression":[155],"is":[156],"effective":[157],"explicit":[160],"retention":[162],"remains":[163],"preferable.":[164]},"counts_by_year":[],"updated_date":"2026-07-29T14:22:42.915294","created_date":"2026-07-02T00:00:00"}