{"id":"https://openalex.org/W4406458080","doi":"https://doi.org/10.1109/bigdata62323.2024.10825615","title":"Problematic Tokens: Tokenizer Bias in Large Language Models","display_name":"Problematic Tokens: Tokenizer Bias in Large Language Models","publication_year":2024,"publication_date":"2024-12-15","ids":{"openalex":"https://openalex.org/W4406458080","doi":"https://doi.org/10.1109/bigdata62323.2024.10825615"},"language":"en","primary_location":{"id":"doi:10.1109/bigdata62323.2024.10825615","is_oa":false,"landing_page_url":"https://doi.org/10.1109/bigdata62323.2024.10825615","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2024 IEEE International Conference on Big Data (BigData)","raw_type":"proceedings-article"},"type":"conference-paper","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/A5102950674","display_name":"Jin Yang","orcid":"https://orcid.org/0009-0009-1706-1444"},"institutions":[{"id":"https://openalex.org/I70983195","display_name":"Syracuse University","ror":"https://ror.org/025r5qe02","country_code":"US","type":"education","lineage":["https://openalex.org/I70983195"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Jin Yang","raw_affiliation_strings":["Syracuse University,Syracuse,USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Syracuse University,Syracuse,USA","institution_ids":["https://openalex.org/I70983195"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5111703907","display_name":"Zhiqiang Wang","orcid":"https://orcid.org/0009-0009-2510-1522"},"institutions":[{"id":"https://openalex.org/I63772739","display_name":"Florida Atlantic University","ror":"https://ror.org/05p8w6387","country_code":"US","type":"education","lineage":["https://openalex.org/I63772739"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Zhiqiang Wang","raw_affiliation_strings":["Florida Atlantic University,Boca Raton,USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Florida Atlantic University,Boca Raton,USA","institution_ids":["https://openalex.org/I63772739"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5109257190","display_name":"Yanbin Lin","orcid":null},"institutions":[{"id":"https://openalex.org/I63772739","display_name":"Florida Atlantic University","ror":"https://ror.org/05p8w6387","country_code":"US","type":"education","lineage":["https://openalex.org/I63772739"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Yanbin Lin","raw_affiliation_strings":["Florida Atlantic University,Boca Raton,USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Florida Atlantic University,Boca Raton,USA","institution_ids":["https://openalex.org/I63772739"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5034112263","display_name":"Zunduo Zhao","orcid":null},"institutions":[{"id":"https://openalex.org/I57206974","display_name":"New York University","ror":"https://ror.org/0190ak572","country_code":"US","type":"education","lineage":["https://openalex.org/I57206974"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Zunduo Zhao","raw_affiliation_strings":["New York University,NYC,USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"New York University,NYC,USA","institution_ids":["https://openalex.org/I57206974"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":3,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":false,"cited_by_count":4,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"6387","last_page":"6393"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10181","display_name":"Natural Language Processing Techniques","score":0.9882000088691711,"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":0.9882000088691711,"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":0.9682000279426575,"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":0.7785465717315674},{"id":"https://openalex.org/keywords/natural-language-processing","display_name":"Natural language processing","score":0.4423346519470215},{"id":"https://openalex.org/keywords/language-model","display_name":"Language model","score":0.4108324646949768}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7785465717315674},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.4423346519470215},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.4108324646949768}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/bigdata62323.2024.10825615","is_oa":false,"landing_page_url":"https://doi.org/10.1109/bigdata62323.2024.10825615","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2024 IEEE International Conference on Big Data (BigData)","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[{"score":0.5099999904632568,"id":"https://metadata.un.org/sdg/10","display_name":"Reduced inequalities"}],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":8,"referenced_works":["https://openalex.org/W4309674289","https://openalex.org/W4392118239","https://openalex.org/W4399116288","https://openalex.org/W4399732199","https://openalex.org/W4404534210","https://openalex.org/W4404792970","https://openalex.org/W6862190762","https://openalex.org/W6868630582"],"related_works":["https://openalex.org/W4391375266","https://openalex.org/W2899084033","https://openalex.org/W2748952813","https://openalex.org/W2390279801","https://openalex.org/W4391913857","https://openalex.org/W2358668433","https://openalex.org/W4396701345","https://openalex.org/W2376932109","https://openalex.org/W2001405890","https://openalex.org/W4396696052"],"abstract_inverted_index":{"Recent":[0],"advancements":[1],"in":[2,17,103],"large":[3],"language":[4],"models":[5,34],"(LLMs),":[6],"such":[7,41],"as":[8,42],"GPT-4":[9],"and":[10,44,84,114,122,143,151,170,176],"GPT-4o,":[11,133],"have":[12],"shown":[13],"exceptional":[14],"performance,":[15],"especially":[16],"languages":[18,40],"with":[19],"abundant":[20],"resources":[21],"like":[22],"English,":[23],"thanks":[24],"to":[25,61,66,78,119,127,147,162,166],"extensive":[26],"datasets":[27],"that":[28],"ensure":[29],"robust":[30],"training.":[31],"Conversely,":[32],"these":[33,59,67,141],"exhibit":[35],"limitations":[36],"when":[37],"processing":[38],"under-resourced":[39],"Chinese":[43],"Korean,":[45],"where":[46],"issues":[47],"including":[48],"hallucinatory":[49],"responses":[50],"remain":[51],"prevalent.":[52],"This":[53,100],"paper":[54],"traces":[55],"the":[56,62,73,81,91,104,129,159],"roots":[57],"of":[58,90,106,132],"disparities":[60],"tokenization":[63,82,130,164],"process":[64,83],"inherent":[65],"models.":[68],"Specifically,":[69],"it":[70],"explores":[71],"how":[72,135],"tokenizer\u2019s":[74],"vocabulary,":[75],"often":[76],"used":[77],"speed":[79],"up":[80],"reduce":[85],"tokens":[86],"but":[87],"constructed":[88],"independently":[89],"actual":[92],"model":[93],"training":[94],"data,":[95],"inadequately":[96],"represents":[97],"non-English":[98],"languages.":[99],"misrepresentation":[101],"results":[102],"propagation":[105],"\u2018under-trained\u2019":[107],"or":[108],"\u2018untrained\u2019":[109],"tokens,":[110],"which":[111],"perpetuate":[112],"biases":[113],"pose":[115],"serious":[116],"concerns":[117],"related":[118],"data":[120,177],"security":[121,150],"ethical":[123,152],"standards.":[124],"We":[125],"aim":[126],"dissect":[128],"mechanics":[131],"illustrating":[134],"its":[136],"simplified":[137],"token-handling":[138],"methods":[139],"amplify":[140],"risks":[142],"offer":[144],"strategic":[145],"solutions":[146],"mitigate":[148],"associated":[149],"issues.":[153],"Through":[154],"this":[155],"study,":[156],"we":[157],"emphasize":[158],"critical":[160],"need":[161],"rethink":[163],"frameworks":[165],"foster":[167],"more":[168],"equitable":[169],"secure":[171],"AI":[172],"technologies.":[173],"The":[174],"code":[175],"are":[178],"available":[179],"at:":[180],"https://github.com/yeyimilk/LLMGPT4o":[181]},"counts_by_year":[{"year":2026,"cited_by_count":2},{"year":2025,"cited_by_count":2}],"updated_date":"2026-07-14T23:27:15.235271","created_date":"2025-10-10T00:00:00"}