{"id":"https://openalex.org/W7140126971","doi":"https://doi.org/10.18653/v1/2026.eacl-long.119","title":"AdaptBPE: From General Purpose to Specialized Tokenizers","display_name":"AdaptBPE: From General Purpose to Specialized Tokenizers","publication_year":2026,"publication_date":"2026-01-01","ids":{"openalex":"https://openalex.org/W7140126971","doi":"https://doi.org/10.18653/v1/2026.eacl-long.119"},"language":"en","primary_location":{"id":"doi:10.18653/v1/2026.eacl-long.119","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.eacl-long.119","pdf_url":"https://aclanthology.org/2026.eacl-long.119.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 19th Conference of the European Chapter of the Association for Computational Linguistics (Volume 1: Long Papers)","raw_type":"proceedings-article"},"type":"article","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://aclanthology.org/2026.eacl-long.119.pdf","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":null,"display_name":"Vijini Pilana Liyanage","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Vijini Pilana Liyanage","raw_affiliation_strings":["Sorbonne-Universit , CNRS , ISIR , Paris"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Sorbonne-Universit , CNRS , ISIR , Paris","institution_ids":[]}]},{"author_position":"last","author":{"id":null,"display_name":"Fran\u00e7ois Yvon","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Fran\u00e7ois Yvon","raw_affiliation_strings":["Sorbonne-Universit , CNRS , ISIR , Paris"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Sorbonne-Universit , CNRS , ISIR , Paris","institution_ids":[]}]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":2,"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":0.47023711,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"2607","last_page":"2620"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10054","display_name":"Parallel Computing and Optimization Techniques","score":0.08730000257492065,"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":0.08730000257492065,"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/T11450","display_name":"Model-Driven Software Engineering Techniques","score":0.08309999853372574,"subfield":{"id":"https://openalex.org/subfields/1712","display_name":"Software"},"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/T10126","display_name":"Logic, programming, and type systems","score":0.07370000332593918,"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/general-purpose","display_name":"General purpose","score":0.37130001187324524},{"id":"https://openalex.org/keywords/general-practice","display_name":"General practice","score":0.3504999876022339},{"id":"https://openalex.org/keywords/key","display_name":"Key (lock)","score":0.24289999902248383},{"id":"https://openalex.org/keywords/term","display_name":"Term (time)","score":0.23579999804496765},{"id":"https://openalex.org/keywords/work","display_name":"Work (physics)","score":0.210999995470047}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.3880999982357025},{"id":"https://openalex.org/C2982832238","wikidata":"https://www.wikidata.org/wiki/Q5531640","display_name":"General purpose","level":2,"score":0.37130001187324524},{"id":"https://openalex.org/C2781318095","wikidata":"https://www.wikidata.org/wiki/Q33127638","display_name":"General practice","level":2,"score":0.3504999876022339},{"id":"https://openalex.org/C127413603","wikidata":"https://www.wikidata.org/wiki/Q11023","display_name":"Engineering","level":0,"score":0.32359999418258667},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.24289999902248383},{"id":"https://openalex.org/C112930515","wikidata":"https://www.wikidata.org/wiki/Q4389547","display_name":"Risk analysis (engineering)","level":1,"score":0.23759999871253967},{"id":"https://openalex.org/C61797465","wikidata":"https://www.wikidata.org/wiki/Q1188986","display_name":"Term (time)","level":2,"score":0.23579999804496765},{"id":"https://openalex.org/C18762648","wikidata":"https://www.wikidata.org/wiki/Q42213","display_name":"Work (physics)","level":2,"score":0.210999995470047},{"id":"https://openalex.org/C9652623","wikidata":"https://www.wikidata.org/wiki/Q190109","display_name":"Field (mathematics)","level":2,"score":0.2093999981880188},{"id":"https://openalex.org/C2777212361","wikidata":"https://www.wikidata.org/wiki/Q5127848","display_name":"Class (philosophy)","level":2,"score":0.20569999516010284}],"mesh":[],"locations_count":2,"locations":[{"id":"doi:10.18653/v1/2026.eacl-long.119","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.eacl-long.119","pdf_url":"https://aclanthology.org/2026.eacl-long.119.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 19th Conference of the European Chapter of the Association for Computational Linguistics (Volume 1: Long Papers)","raw_type":"proceedings-article"},{"id":"pmh:oai:HAL:hal-05563266v1","is_oa":true,"landing_page_url":"https://hal.science/hal-05563266","pdf_url":"https://hal.science/hal-05563266v1/file/2026.eacl-long.119.pdf","source":{"id":"https://openalex.org/S4306402512","display_name":"HAL (Le Centre pour la Communication Scientifique Directe)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I1294671590","host_organization_name":"Centre National de la Recherche Scientifique","host_organization_lineage":["https://openalex.org/I1294671590"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by-nc-nd","license_id":"https://openalex.org/licenses/cc-by-nc-nd","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"19th Conference of the European Chapter of the Association for Computational Linguistics, Association for Computational Linguistics, Mar 2026, Rabat, Morocco. pp.2607--2620, ⟨10.18653/v1/2026.eacl-long.119⟩","raw_type":"Conference papers"}],"best_oa_location":{"id":"doi:10.18653/v1/2026.eacl-long.119","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.eacl-long.119","pdf_url":"https://aclanthology.org/2026.eacl-long.119.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 19th Conference of the European Chapter of the Association for Computational Linguistics (Volume 1: Long Papers)","raw_type":"proceedings-article"},"sustainable_development_goals":[],"awards":[{"id":"https://openalex.org/G613589555","display_name":"Translating with Large Language Models","funder_award_id":"ANR-23-IAS1-0006","funder_id":"https://openalex.org/F4320320883","funder_display_name":"Agence Nationale de la Recherche"}],"funders":[{"id":"https://openalex.org/F4320320883","display_name":"Agence Nationale de la Recherche","ror":"https://ror.org/00rbzpz17"},{"id":"https://openalex.org/F4320326256","display_name":"Grand \u00c9quipement National De Calcul Intensif","ror":"https://ror.org/0010d1q40"}],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7140126971.pdf","grobid_xml":"https://content.openalex.org/works/W7140126971.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Subword":[0],"tokenization":[1,142],"methods,":[2],"such":[3],"as":[4,129],"Byte-Pair":[5],"Encoding":[6],"(BPE),":[7],"significantly":[8],"impact":[9],"the":[10,37,50,85,92,123],"performance":[11],"and":[12,35,104,149],"efficiency":[13],"of":[14,39,43],"large":[15],"language":[16],"models":[17],"(LLMs).The":[18],"standard":[19],"approach":[20],"involves":[21],"training":[22,34],"a":[23,40,62,96,130,136],"general-purpose":[24],"tokenizer":[25],"that":[26,66,88,111],"uniformly":[27],"processes":[28],"all":[29],"textual":[30],"data":[31,150],"during":[32],"both":[33],"inference.However,":[36],"use":[38],"generic":[41],"set":[42],"tokens":[44,70],"can":[45],"incur":[46],"inefficiencies":[47],"when":[48],"applying":[49],"model":[51],"to":[52,135],"specific":[53,144],"domains":[54,145],"or":[55,146],"languages.To":[56],"address":[57],"this":[58],"limitation,":[59],"we":[60],"propose":[61],"post-training":[63],"adaptation":[64,81,93,132],"strategy":[65],"selectively":[67],"replaces":[68],"lowutility":[69],"with":[71],"more":[72,118],"relevant":[73],"ones":[74],"based":[75],"on":[76,102],"their":[77],"frequency":[78],"in":[79],"an":[80],"corpus.Our":[82],"algorithm":[83],"identifies":[84],"token":[86],"inventory":[87],"most":[89],"effectively":[90,119],"encodes":[91],"corpus":[94],"for":[95,143],"given":[97],"target":[98],"vocabulary":[99,125,137],"size.Extensive":[100],"experiments":[101],"generation":[103],"classification":[105],"tasks":[106],"across":[107],"multiple":[108],"languages":[109],"demonstrate":[110],"our":[112],"adapted":[113],"tokenizers":[114],"compress":[115],"test":[116],"corpora":[117],"than":[120],"baselines":[121],"using":[122],"same":[124],"size.This":[126],"method":[127],"serves":[128],"lightweight":[131],"mechanism,":[133],"akin":[134],"fine-tuning":[138],"process,":[139],"enabling":[140],"optimized":[141],"tasks.Our":[147],"code":[148],"are":[151],"available":[152],"at":[153],"https:":[154],"//github.com/vijini/Adapt-BPE.git.":[155]},"counts_by_year":[],"updated_date":"2026-06-11T09:08:48.828518","created_date":"2026-02-02T00:00:00"}