{"id":"https://openalex.org/W4408352241","doi":"https://doi.org/10.1109/icassp49660.2025.10889733","title":"LBPE: Long-token-first Tokenization to Improve Large Language Models","display_name":"LBPE: Long-token-first Tokenization to Improve Large Language Models","publication_year":2025,"publication_date":"2025-03-12","ids":{"openalex":"https://openalex.org/W4408352241","doi":"https://doi.org/10.1109/icassp49660.2025.10889733"},"language":"en","primary_location":{"id":"doi:10.1109/icassp49660.2025.10889733","is_oa":false,"landing_page_url":"https://doi.org/10.1109/icassp49660.2025.10889733","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"ICASSP 2025 - 2025 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)","raw_type":"proceedings-article"},"type":"article","indexed_in":["crossref"],"open_access":{"is_oa":false,"oa_status":"closed","oa_url":null,"any_repository_has_fulltext":false},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5060527527","display_name":"Haoran Lian","orcid":null},"institutions":[{"id":"https://openalex.org/I82880672","display_name":"Beihang University","ror":"https://ror.org/00wk2mp56","country_code":"CN","type":"education","lineage":["https://openalex.org/I82880672"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Haoran Lian","raw_affiliation_strings":["Beihang University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Beihang University","institution_ids":["https://openalex.org/I82880672"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101351925","display_name":"Yizhe Xiong","orcid":null},"institutions":[{"id":"https://openalex.org/I99065089","display_name":"Tsinghua University","ror":"https://ror.org/03cve4549","country_code":"CN","type":"education","lineage":["https://openalex.org/I99065089"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Yizhe Xiong","raw_affiliation_strings":["Tsinghua University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Tsinghua University","institution_ids":["https://openalex.org/I99065089"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5077882811","display_name":"Zijia Lin","orcid":"https://orcid.org/0000-0002-1390-7424"},"institutions":[{"id":"https://openalex.org/I99065089","display_name":"Tsinghua University","ror":"https://ror.org/03cve4549","country_code":"CN","type":"education","lineage":["https://openalex.org/I99065089"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Zijia Lin","raw_affiliation_strings":["Tsinghua University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Tsinghua University","institution_ids":["https://openalex.org/I99065089"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5053316728","display_name":"Jianwei Niu","orcid":"https://orcid.org/0000-0003-3946-5107"},"institutions":[{"id":"https://openalex.org/I82880672","display_name":"Beihang University","ror":"https://ror.org/00wk2mp56","country_code":"CN","type":"education","lineage":["https://openalex.org/I82880672"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Jianwei Niu","raw_affiliation_strings":["Beihang University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Beihang University","institution_ids":["https://openalex.org/I82880672"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5078214911","display_name":"Shasha Mo","orcid":null},"institutions":[{"id":"https://openalex.org/I82880672","display_name":"Beihang University","ror":"https://ror.org/00wk2mp56","country_code":"CN","type":"education","lineage":["https://openalex.org/I82880672"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Shasha Mo","raw_affiliation_strings":["Beihang University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Beihang University","institution_ids":["https://openalex.org/I82880672"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5113744795","display_name":"Hui Chen","orcid":"https://orcid.org/0000-0001-6647-1137"},"institutions":[{"id":"https://openalex.org/I99065089","display_name":"Tsinghua University","ror":"https://ror.org/03cve4549","country_code":"CN","type":"education","lineage":["https://openalex.org/I99065089"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Hui Chen","raw_affiliation_strings":["Tsinghua University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Tsinghua University","institution_ids":["https://openalex.org/I99065089"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100346758","display_name":"Peng Liu","orcid":"https://orcid.org/0000-0001-7518-1713"},"institutions":[{"id":"https://openalex.org/I125839683","display_name":"Beijing Institute of Technology","ror":"https://ror.org/01skt4w74","country_code":"CN","type":"education","lineage":["https://openalex.org/I125839683","https://openalex.org/I890469752"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Peng Liu","raw_affiliation_strings":["Beijing Institute of Technology"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Beijing Institute of Technology","institution_ids":["https://openalex.org/I125839683"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5057732142","display_name":"Guiguang Ding","orcid":"https://orcid.org/0000-0003-0137-9975"},"institutions":[{"id":"https://openalex.org/I99065089","display_name":"Tsinghua University","ror":"https://ror.org/03cve4549","country_code":"CN","type":"education","lineage":["https://openalex.org/I99065089"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Guiguang Ding","raw_affiliation_strings":["Tsinghua University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Tsinghua University","institution_ids":["https://openalex.org/I99065089"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":8,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":1.7588,"has_fulltext":false,"cited_by_count":1,"citation_normalized_percentile":{"value":0.85561784,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":91,"max":95},"biblio":{"volume":null,"issue":null,"first_page":"1","last_page":"5"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10181","display_name":"Natural Language Processing Techniques","score":0.9815000295639038,"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/T10181","display_name":"Natural Language Processing Techniques","score":0.9815000295639038,"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/T10028","display_name":"Topic Modeling","score":0.9786999821662903,"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/computer-science","display_name":"Computer science","score":0.8081008195877075},{"id":"https://openalex.org/keywords/lexical-analysis","display_name":"Lexical analysis","score":0.8007233738899231},{"id":"https://openalex.org/keywords/security-token","display_name":"Security token","score":0.7049953937530518},{"id":"https://openalex.org/keywords/programming-language","display_name":"Programming language","score":0.34730446338653564},{"id":"https://openalex.org/keywords/natural-language-processing","display_name":"Natural language processing","score":0.3241332471370697},{"id":"https://openalex.org/keywords/computer-network","display_name":"Computer network","score":0.20809954404830933}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8081008195877075},{"id":"https://openalex.org/C176982825","wikidata":"https://www.wikidata.org/wiki/Q835922","display_name":"Lexical analysis","level":2,"score":0.8007233738899231},{"id":"https://openalex.org/C48145219","wikidata":"https://www.wikidata.org/wiki/Q1335365","display_name":"Security token","level":2,"score":0.7049953937530518},{"id":"https://openalex.org/C199360897","wikidata":"https://www.wikidata.org/wiki/Q9143","display_name":"Programming language","level":1,"score":0.34730446338653564},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.3241332471370697},{"id":"https://openalex.org/C31258907","wikidata":"https://www.wikidata.org/wiki/Q1301371","display_name":"Computer network","level":1,"score":0.20809954404830933}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/icassp49660.2025.10889733","is_oa":false,"landing_page_url":"https://doi.org/10.1109/icassp49660.2025.10889733","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"ICASSP 2025 - 2025 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[],"awards":[],"funders":[{"id":"https://openalex.org/F4320321001","display_name":"National Natural Science Foundation of China","ror":"https://ror.org/01h0zpd94"}],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":28,"referenced_works":["https://openalex.org/W2466175319","https://openalex.org/W2890894339","https://openalex.org/W2946609015","https://openalex.org/W2962784628","https://openalex.org/W2970062726","https://openalex.org/W2998617917","https://openalex.org/W3154046803","https://openalex.org/W3175212568","https://openalex.org/W3176453404","https://openalex.org/W3194676777","https://openalex.org/W4312886719","https://openalex.org/W4385569713","https://openalex.org/W4401042301","https://openalex.org/W4411119384","https://openalex.org/W6601894380","https://openalex.org/W6764401283","https://openalex.org/W6772383348","https://openalex.org/W6778883912","https://openalex.org/W6788175385","https://openalex.org/W6811340617","https://openalex.org/W6850625674","https://openalex.org/W6851579256","https://openalex.org/W6854866820","https://openalex.org/W6856398428","https://openalex.org/W6866362821","https://openalex.org/W6866901960","https://openalex.org/W6870439706","https://openalex.org/W6874724984"],"related_works":["https://openalex.org/W2899084033","https://openalex.org/W2748952813","https://openalex.org/W4298195702","https://openalex.org/W4300598845","https://openalex.org/W2601638452","https://openalex.org/W2945402993","https://openalex.org/W2285263069","https://openalex.org/W4376107815","https://openalex.org/W4319309671","https://openalex.org/W4319309603"],"abstract_inverted_index":{"The":[0],"prevalent":[1],"use":[2],"of":[3,16,22,80],"Byte":[4],"Pair":[5],"Encoding":[6],"(BPE)":[7],"in":[8,35,41,51,87],"Large":[9],"Language":[10],"Models":[11],"(LLMs)":[12],"facilitates":[13],"robust":[14],"handling":[15],"subword":[17],"units":[18],"and":[19,107,110],"avoids":[20],"issues":[21],"out-of-vocabulary":[23],"words.":[24],"Despite":[25],"its":[26,133],"success,":[27],"a":[28],"critical":[29],"challenge":[30],"persists:":[31],"long":[32,66,108],"tokens,":[33,47,109],"rich":[34],"semantic":[36],"information,":[37],"have":[38],"fewer":[39],"occurrences":[40],"tokenized":[42],"datasets":[43],"compared":[44],"to":[45,76],"short":[46,106],"which":[48,64],"can":[49],"result":[50],"imbalanced":[52],"learning":[53,114],"issue":[54],"across":[55,118],"different":[56],"tokens.":[57],"To":[58],"address":[59],"that,":[60],"we":[61],"propose":[62],"LBPE,":[63],"prioritizes":[65],"tokens":[67,74,92],"during":[68,95],"the":[69,88,96,102,113,128],"encoding":[70,97],"process.":[71,98],"LBPE":[72,100,125],"generates":[73],"according":[75],"their":[77,85],"descending":[78],"order":[79],"token":[81],"length":[82],"rather":[83],"than":[84],"ranks":[86],"vocabulary,":[89],"granting":[90],"longer":[91],"higher":[93],"priority":[94],"Consequently,":[99],"smooths":[101],"frequency":[103],"differences":[104],"between":[105],"thus":[111],"mitigates":[112],"imbalance.":[115],"Extensive":[116],"experiments":[117],"diverse":[119],"language":[120],"modeling":[121],"tasks":[122],"demonstrate":[123],"that":[124],"consistently":[126],"outperforms":[127],"original":[129],"BPE,":[130],"well":[131],"demonstrating":[132],"effectiveness.":[134]},"counts_by_year":[{"year":2025,"cited_by_count":1}],"updated_date":"2026-06-11T09:08:48.828518","created_date":"2025-10-10T00:00:00"}