{"id":"https://openalex.org/W7161161936","doi":"https://doi.org/10.48550/arxiv.2605.12928","title":"The Efficiency Gap in Byte Modeling","display_name":"The Efficiency Gap in Byte Modeling","publication_year":2026,"publication_date":"2026-05-13","ids":{"openalex":"https://openalex.org/W7161161936","doi":"https://doi.org/10.48550/arxiv.2605.12928"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2605.12928","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.12928","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"},"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.48550/arxiv.2605.12928","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5136090835","display_name":"Celine Lee","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Lee, Celine","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5113817523","display_name":"Jing Nathan Yan","orcid":"https://orcid.org/0009-0004-0157-5065"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yan, Jing Nathan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5136137610","display_name":"Chen Liang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Liang, Chen","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5136159912","display_name":"Jiaxin Shi","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Shi, Jiaxin","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5136156855","display_name":"Yin Zhang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhang, Yin","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5054327137","display_name":"Jeremiah Liu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Liu, Jeremiah","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5136121966","display_name":"Pengcheng Yin","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yin, Pengcheng","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5125228889","display_name":"Fernando Pereira","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Pereira, Fernando","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5136131917","display_name":"Ed Chi","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Chi, Ed","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5127816692","display_name":"Derek Cheng","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Cheng, Derek","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5062546146","display_name":"Alexander M. Rush","orcid":"https://orcid.org/0000-0002-9900-1606"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Rush, Alexander M.","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5136141195","display_name":"Ruoxi Wang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wang, Ruoxi","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":false,"cited_by_count":0,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":null,"last_page":null},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10028","display_name":"Topic Modeling","score":0.6146000027656555,"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/T10028","display_name":"Topic Modeling","score":0.6146000027656555,"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/T13910","display_name":"Computational and Text Analysis Methods","score":0.05510000139474869,"subfield":{"id":"https://openalex.org/subfields/3300","display_name":"General Social Sciences"},"field":{"id":"https://openalex.org/fields/33","display_name":"Social Sciences"},"domain":{"id":"https://openalex.org/domains/2","display_name":"Social Sciences"}},{"id":"https://openalex.org/T10181","display_name":"Natural Language Processing Techniques","score":0.05350000038743019,"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/byte","display_name":"Byte","score":0.6241999864578247},{"id":"https://openalex.org/keywords/context","display_name":"Context (archaeology)","score":0.5443000197410583},{"id":"https://openalex.org/keywords/prior-probability","display_name":"Prior probability","score":0.4837999939918518},{"id":"https://openalex.org/keywords/intersection","display_name":"Intersection (aeronautics)","score":0.44839999079704285},{"id":"https://openalex.org/keywords/scaling","display_name":"Scaling","score":0.44600000977516174},{"id":"https://openalex.org/keywords/language-model","display_name":"Language model","score":0.4147000014781952},{"id":"https://openalex.org/keywords/spurious-relationship","display_name":"Spurious relationship","score":0.4025000035762787},{"id":"https://openalex.org/keywords/overhead","display_name":"Overhead (engineering)","score":0.3720000088214874}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7297999858856201},{"id":"https://openalex.org/C43364308","wikidata":"https://www.wikidata.org/wiki/Q8799","display_name":"Byte","level":2,"score":0.6241999864578247},{"id":"https://openalex.org/C2779343474","wikidata":"https://www.wikidata.org/wiki/Q3109175","display_name":"Context (archaeology)","level":2,"score":0.5443000197410583},{"id":"https://openalex.org/C177769412","wikidata":"https://www.wikidata.org/wiki/Q278090","display_name":"Prior probability","level":3,"score":0.4837999939918518},{"id":"https://openalex.org/C64543145","wikidata":"https://www.wikidata.org/wiki/Q162942","display_name":"Intersection (aeronautics)","level":2,"score":0.44839999079704285},{"id":"https://openalex.org/C99844830","wikidata":"https://www.wikidata.org/wiki/Q102441924","display_name":"Scaling","level":2,"score":0.44600000977516174},{"id":"https://openalex.org/C80444323","wikidata":"https://www.wikidata.org/wiki/Q2878974","display_name":"Theoretical computer science","level":1,"score":0.41499999165534973},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.4147000014781952},{"id":"https://openalex.org/C97256817","wikidata":"https://www.wikidata.org/wiki/Q1462316","display_name":"Spurious relationship","level":2,"score":0.4025000035762787},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.37619999051094055},{"id":"https://openalex.org/C2779960059","wikidata":"https://www.wikidata.org/wiki/Q7113681","display_name":"Overhead (engineering)","level":2,"score":0.3720000088214874},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.3637999892234802},{"id":"https://openalex.org/C184337299","wikidata":"https://www.wikidata.org/wiki/Q1437428","display_name":"Semantics (computer science)","level":2,"score":0.3456000089645386},{"id":"https://openalex.org/C183322885","wikidata":"https://www.wikidata.org/wiki/Q17007702","display_name":"Context model","level":3,"score":0.3246000111103058},{"id":"https://openalex.org/C21308566","wikidata":"https://www.wikidata.org/wiki/Q7169365","display_name":"Permutation (music)","level":2,"score":0.31610000133514404},{"id":"https://openalex.org/C48145219","wikidata":"https://www.wikidata.org/wiki/Q1335365","display_name":"Security token","level":2,"score":0.3091999888420105},{"id":"https://openalex.org/C107673813","wikidata":"https://www.wikidata.org/wiki/Q812534","display_name":"Bayesian probability","level":2,"score":0.3073999881744385},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.3001999855041504},{"id":"https://openalex.org/C31170391","wikidata":"https://www.wikidata.org/wiki/Q188619","display_name":"Hierarchy","level":2,"score":0.29829999804496765},{"id":"https://openalex.org/C159877910","wikidata":"https://www.wikidata.org/wiki/Q2202883","display_name":"Autoregressive model","level":2,"score":0.27619999647140503},{"id":"https://openalex.org/C125411270","wikidata":"https://www.wikidata.org/wiki/Q18653","display_name":"Encoding (memory)","level":2,"score":0.2680000066757202},{"id":"https://openalex.org/C114466953","wikidata":"https://www.wikidata.org/wiki/Q6034165","display_name":"Initialization","level":2,"score":0.26190000772476196},{"id":"https://openalex.org/C160234255","wikidata":"https://www.wikidata.org/wiki/Q812535","display_name":"Bayesian inference","level":3,"score":0.2615000009536743},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.2590999901294708},{"id":"https://openalex.org/C99498987","wikidata":"https://www.wikidata.org/wiki/Q2210247","display_name":"Noise (video)","level":3,"score":0.25679999589920044}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2605.12928","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.12928","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2605.12928","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.12928","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"},"sustainable_development_goals":[{"score":0.4457128345966339,"id":"https://metadata.un.org/sdg/4","display_name":"Quality Education"}],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Modern":[0],"language":[1],"models":[2,127],"have":[3,32],"historically":[4],"relied":[5],"on":[6],"two":[7,29],"dominant":[8],"design":[9,18],"choices:":[10],"subword":[11,131],"tokenization":[12],"and":[13,43],"autoregressive":[14],"(AR)":[15],"ordering.":[16],"These":[17],"decisions":[19],"bake":[20],"in":[21,169],"priors":[22,66],"that":[23,87,114,155],"dictate":[24],"a":[25,56,68,80],"model's":[26],"learning.":[27],"Recently,":[28],"alternative":[30,161],"paradigms":[31],"challenged":[33],"this:":[34],"byte-level":[35],"modeling,":[36],"which":[37,48],"bypasses":[38],"static":[39],"statistically-derived":[40],"token":[41],"vocabularies,":[42],"masked":[44],"diffusion":[45],"modeling":[46,93,104],"(MDM),":[47],"conducts":[49],"parallel,":[50],"non-sequential":[51],"generation.":[52],"Their":[53],"intersection":[54],"represents":[55],"fully":[57],"end-to-end":[58],"modality-agnostic":[59,157],"generative":[60],"prototype;":[61],"however,":[62],"removing":[63],"these":[64],"structural":[65,162],"incurs":[67],"significant":[69],"computational":[70],"cost.":[71],"In":[72],"this":[73,77,115],"work,":[74],"we":[75],"investigate":[76],"cost":[78],"through":[79],"compute-matched":[81],"scaling":[82,100,167],"study.":[83],"Our":[84,148],"results":[85],"reveal":[86],"the":[88,99,133,137,170],"performance":[89],"penalty":[90],"of":[91,102],"byte":[92,103,171],"is":[94,105],"not":[95],"uniform;":[96],"across":[97],"scale,":[98],"overhead":[101],"worse":[106],"for":[107,110],"MDM":[108,134],"than":[109],"AR.":[111],"We":[112],"hypothesize":[113],"disparity":[116],"stems":[117],"from":[118,145,150],"context":[119],"fragility:":[120],"while":[121],"AR's":[122],"stable":[123],"causal":[124],"history":[125],"allows":[126],"to":[128,141,164],"naturally":[129],"rediscover":[130],"patterns,":[132],"objective":[135],"destroys":[136],"local":[138],"contiguity":[139],"required":[140],"efficiently":[142],"resolve":[143],"semantics":[144],"raw":[146],"bytes.":[147],"findings":[149],"controlled":[151],"permutation":[152],"experiments":[153],"suggest":[154],"future":[156],"designs":[158],"must":[159],"incorporate":[160],"biases":[163],"maintain":[165],"viable":[166],"trajectories":[168],"regime.":[172]},"counts_by_year":[],"updated_date":"2026-07-01T06:00:48.157686","created_date":"2026-05-15T00:00:00"}