{"id":"https://openalex.org/W7140132448","doi":"https://doi.org/10.18653/v1/2026.findings-eacl.181","title":"DWA-KD: Dual-Space Weighting and Time-Warped Alignment for Cross-Tokenizer Knowledge Distillation","display_name":"DWA-KD: Dual-Space Weighting and Time-Warped Alignment for Cross-Tokenizer Knowledge Distillation","publication_year":2026,"publication_date":"2026-01-01","ids":{"openalex":"https://openalex.org/W7140132448","doi":"https://doi.org/10.18653/v1/2026.findings-eacl.181"},"language":null,"primary_location":{"id":"doi:10.18653/v1/2026.findings-eacl.181","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.findings-eacl.181","pdf_url":"https://aclanthology.org/2026.findings-eacl.181.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: EACL 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-eacl.181.pdf","any_repository_has_fulltext":null},"authorships":[{"author_position":"first","author":{"id":null,"display_name":"Duc Trung Vu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Duc Trung Vu","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":null,"display_name":"Pham Khanh Chi","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Pham Khanh Chi","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":null,"display_name":"Dat Phi Van","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Dat Phi Van","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":null,"display_name":"Linh Ngo Van","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Linh Ngo Van","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":null,"display_name":"Dinh Viet Sang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Dinh Viet Sang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":null,"display_name":"Trung Le","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Trung Le","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":true,"cited_by_count":0,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"3513","last_page":"3527"},"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":0.07429999858140945,"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":0.07429999858140945,"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/T11273","display_name":"Advanced Graph Neural Networks","score":0.07209999859333038,"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/T10036","display_name":"Advanced Neural Network Applications","score":0.06610000133514404,"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/weighting","display_name":"Weighting","score":0.5418000221252441},{"id":"https://openalex.org/keywords/pattern-recognition","display_name":"Pattern recognition (psychology)","score":0.27959999442100525},{"id":"https://openalex.org/keywords/expert-system","display_name":"Expert system","score":0.2727000117301941},{"id":"https://openalex.org/keywords/feature","display_name":"Feature (linguistics)","score":0.27160000801086426},{"id":"https://openalex.org/keywords/distillation","display_name":"Distillation","score":0.26190000772476196}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.5471000075340271},{"id":"https://openalex.org/C183115368","wikidata":"https://www.wikidata.org/wiki/Q856577","display_name":"Weighting","level":2,"score":0.5418000221252441},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.4562999904155731},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.3634999990463257},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.3059000074863434},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.2921999990940094},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.27959999442100525},{"id":"https://openalex.org/C58328972","wikidata":"https://www.wikidata.org/wiki/Q184609","display_name":"Expert system","level":2,"score":0.2727000117301941},{"id":"https://openalex.org/C2776401178","wikidata":"https://www.wikidata.org/wiki/Q12050496","display_name":"Feature (linguistics)","level":2,"score":0.27160000801086426},{"id":"https://openalex.org/C204030448","wikidata":"https://www.wikidata.org/wiki/Q101017","display_name":"Distillation","level":2,"score":0.26190000772476196},{"id":"https://openalex.org/C115925183","wikidata":"https://www.wikidata.org/wiki/Q1412694","display_name":"Knowledge-based systems","level":2,"score":0.26019999384880066}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.18653/v1/2026.findings-eacl.181","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.findings-eacl.181","pdf_url":"https://aclanthology.org/2026.findings-eacl.181.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: EACL 2026","raw_type":"proceedings-article"}],"best_oa_location":{"id":"doi:10.18653/v1/2026.findings-eacl.181","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.findings-eacl.181","pdf_url":"https://aclanthology.org/2026.findings-eacl.181.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: EACL 2026","raw_type":"proceedings-article"},"sustainable_development_goals":[],"awards":[{"id":"https://openalex.org/G7110693004","display_name":null,"funder_award_id":"102.05","funder_id":"https://openalex.org/F4320309617","funder_display_name":"National Foundation for Science and Technology Development"}],"funders":[{"id":"https://openalex.org/F4320309617","display_name":"National Foundation for Science and Technology Development","ror":"https://ror.org/04rw64z44"}],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7140132448.pdf","grobid_xml":"https://content.openalex.org/works/W7140132448.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Knowledge":[0],"Distillation":[1],"(KD)":[2],"has":[3],"emerged":[4],"as":[5],"a":[6,46],"crucial":[7],"technique":[8],"for":[9],"compressing":[10],"Large":[11],"Language":[12],"Models":[13],"(LLMs).Although":[14],"existing":[15],"cross-tokenizer":[16,48],"KD":[17,87,164],"methods":[18],"have":[19],"made":[20],"notable":[21],"progress,":[22],"their":[23],"effectiveness":[24],"remains":[25],"constrained":[26],"by":[27,64,95],"suboptimal":[28],"alignment":[29,63,143],"across":[30,155],"sequence":[31,124],"and":[32,42,59,68,82,106,137,146,151,177,179],"vocabulary":[33],"levels.To":[34],"address":[35],"these":[36],"limitations,":[37],"we":[38],"introduce":[39],"Dual-Space":[40],"Weighting":[41],"Time-Warped":[43],"Alignment":[44],"(DWA-KD),":[45],"novel":[47],"distillation":[49,54],"framework":[50],"that":[51,98,160],"enhances":[52],"token-wise":[53],"through":[55],"dual-space":[56,86,96],"entropy-based":[57,174],"weighting":[58,176],"achieves":[60],"precise":[61],"sequence-level":[62],"leveraging":[65],"both":[66,134],"lexical":[67,145],"semantic":[69],"information.At":[70],"the":[71,79,102,107,123,135,170],"token":[72,175],"level,":[73,125],"DWA-KD":[74,126,161],"maps":[75],"teacher":[76,108,150],"representations":[77],"into":[78],"student":[80,103,152],"space":[81],"vice":[83],"versa,":[84],"performing":[85],"via":[88],"Kullback-Leibler":[89],"divergence":[90],"(KL).The":[91],"process":[92],"is":[93,104,109],"modulated":[94],"weights":[97],"up-weight":[99],"tokens":[100,116],"where":[101],"uncertain":[105],"confident,":[110],"thereby":[111],"focusing":[112],"learning":[113],"on":[114],"informative":[115],"rather":[117],"than":[118],"treating":[119],"all":[120],"positions":[121],"equally.At":[122],"applies":[127],"Soft":[128],"Dynamic":[129],"Time":[130],"Warping":[131],"(Soft-DTW)":[132],"to":[133],"embedding":[136,178],"final":[138,180],"hidden-state":[139],"layers,":[140],"enabling":[141],"robust":[142],"of":[144,173],"contextual":[147],"semantics":[148],"between":[149],"sequences.Extensive":[153],"experiments":[154],"diverse":[156],"NLP":[157],"benchmarks":[158],"demonstrate":[159],"outperforms":[162],"state-of-the-art":[163],"baselines,":[165],"while":[166],"ablation":[167],"studies":[168],"confirm":[169],"complementary":[171],"contributions":[172],"hidden":[181],"state":[182],"layer":[183],"Soft-DTW":[184],"alignment.":[185]},"counts_by_year":[],"updated_date":"2026-07-14T23:27:15.235271","created_date":"2026-03-01T00:00:00"}