{"id":"https://openalex.org/W7123704810","doi":"https://doi.org/10.48550/arxiv.2601.06329","title":"On the Fallacy of Global Token Perplexity in Spoken Language Model Evaluation","display_name":"On the Fallacy of Global Token Perplexity in Spoken Language Model Evaluation","publication_year":2026,"publication_date":"2026-01-09","ids":{"openalex":"https://openalex.org/W7123704810","doi":"https://doi.org/10.48550/arxiv.2601.06329"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2601.06329","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2601.06329","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"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.2601.06329","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5122952826","display_name":"Jeff Chan-Jan Sju","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Hsu, Chan-Jan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5044228068","display_name":"Liang-Hsuan Tseng","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Tseng, Liang-Hsuan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5105414551","display_name":"Yi-Cheng Lin","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Lin, Yi-Cheng","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5122937118","display_name":"Yen-Chun Kuo","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Kuo, Yen-Chun","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5049062211","display_name":"Ju-Chieh Chou","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Chou, Ju-Chieh","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5122925079","display_name":"Kai-Wei Chang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Chang, Kai-Wei","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5122923609","display_name":"Hung-yi Lee","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Lee, Hung-yi","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5040793194","display_name":"Carlos Busso","orcid":"https://orcid.org/0000-0002-4075-4072"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Busso, Carlos","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/T10201","display_name":"Speech Recognition and Synthesis","score":0.2565000057220459,"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/T10201","display_name":"Speech Recognition and Synthesis","score":0.2565000057220459,"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/T11448","display_name":"Face recognition and analysis","score":0.22689999639987946,"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"}},{"id":"https://openalex.org/T10667","display_name":"Emotion and Mood Recognition","score":0.18520000576972961,"subfield":{"id":"https://openalex.org/subfields/3205","display_name":"Experimental and Cognitive Psychology"},"field":{"id":"https://openalex.org/fields/32","display_name":"Psychology"},"domain":{"id":"https://openalex.org/domains/2","display_name":"Social Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/perplexity","display_name":"Perplexity","score":0.9229999780654907},{"id":"https://openalex.org/keywords/spoken-language","display_name":"Spoken language","score":0.7595999836921692},{"id":"https://openalex.org/keywords/language-model","display_name":"Language model","score":0.6028000116348267},{"id":"https://openalex.org/keywords/security-token","display_name":"Security token","score":0.578000009059906},{"id":"https://openalex.org/keywords/variety","display_name":"Variety (cybernetics)","score":0.49729999899864197},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.4652000069618225},{"id":"https://openalex.org/keywords/generative-model","display_name":"Generative model","score":0.4108000099658966}],"concepts":[{"id":"https://openalex.org/C100279451","wikidata":"https://www.wikidata.org/wiki/Q372193","display_name":"Perplexity","level":3,"score":0.9229999780654907},{"id":"https://openalex.org/C2776230583","wikidata":"https://www.wikidata.org/wiki/Q1322198","display_name":"Spoken language","level":2,"score":0.7595999836921692},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7416999936103821},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.6136999726295471},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.6028000116348267},{"id":"https://openalex.org/C48145219","wikidata":"https://www.wikidata.org/wiki/Q1335365","display_name":"Security token","level":2,"score":0.578000009059906},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.531499981880188},{"id":"https://openalex.org/C136197465","wikidata":"https://www.wikidata.org/wiki/Q1729295","display_name":"Variety (cybernetics)","level":2,"score":0.49729999899864197},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.4652000069618225},{"id":"https://openalex.org/C28490314","wikidata":"https://www.wikidata.org/wiki/Q189436","display_name":"Speech recognition","level":1,"score":0.42179998755455017},{"id":"https://openalex.org/C167966045","wikidata":"https://www.wikidata.org/wiki/Q5532625","display_name":"Generative model","level":3,"score":0.4108000099658966},{"id":"https://openalex.org/C39890363","wikidata":"https://www.wikidata.org/wiki/Q36108","display_name":"Generative grammar","level":2,"score":0.40299999713897705},{"id":"https://openalex.org/C2781035248","wikidata":"https://www.wikidata.org/wiki/Q186150","display_name":"Fallacy","level":2,"score":0.3634999990463257},{"id":"https://openalex.org/C41895202","wikidata":"https://www.wikidata.org/wiki/Q8162","display_name":"Linguistics","level":1,"score":0.30970001220703125},{"id":"https://openalex.org/C61328038","wikidata":"https://www.wikidata.org/wiki/Q3358061","display_name":"Speech processing","level":2,"score":0.28679999709129333},{"id":"https://openalex.org/C195324797","wikidata":"https://www.wikidata.org/wiki/Q33742","display_name":"Natural language","level":2,"score":0.27489998936653137},{"id":"https://openalex.org/C2780844864","wikidata":"https://www.wikidata.org/wiki/Q184377","display_name":"Pronunciation","level":2,"score":0.27230000495910645},{"id":"https://openalex.org/C74672266","wikidata":"https://www.wikidata.org/wiki/Q815859","display_name":"Language acquisition","level":2,"score":0.2596000134944916},{"id":"https://openalex.org/C15744967","wikidata":"https://www.wikidata.org/wiki/Q9418","display_name":"Psychology","level":0,"score":0.25060001015663147}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2601.06329","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2601.06329","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"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.2601.06329","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2601.06329","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"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":[],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Generative":[0],"spoken":[1,29,129,161],"language":[2,130,162],"models":[3,27,35,131],"pretrained":[4],"on":[5],"large-scale":[6],"raw":[7],"audio":[8],"can":[9],"continue":[10],"a":[11,78,135],"speech":[12,51,60,71],"prompt":[13],"with":[14,112],"appropriate":[15,152],"content":[16],"while":[17],"preserving":[18],"attributes":[19],"like":[20],"speaker":[21],"and":[22,61,82,143],"emotion,":[23],"serving":[24],"as":[25,107],"foundation":[26],"for":[28,156],"dialogue.":[30],"In":[31,73],"prior":[32],"literature,":[33],"these":[34,148],"are":[36],"often":[37],"evaluated":[38],"using":[39],"``global":[40],"token":[41,93],"perplexity'',":[42],"which":[43],"directly":[44],"applies":[45],"the":[46,70,98,121,124,140,144],"text":[47,62],"perplexity":[48],"formulation":[49],"to":[50,66],"tokens.":[52],"However,":[53],"this":[54,74],"practice":[55],"overlooks":[56],"fundamental":[57],"differences":[58],"between":[59,139],"modalities,":[63],"possibly":[64],"leading":[65],"an":[67],"underestimation":[68],"of":[69,80,90,128],"characteristics.":[72],"work,":[75],"we":[76],"propose":[77],"variety":[79],"likelihood-":[81],"generative-based":[83],"evaluation":[84,153],"methods":[85],"that":[86,97,151],"serve":[87],"in":[88,160],"place":[89],"naive":[91],"global":[92],"perplexity.":[94],"We":[95],"demonstrate":[96],"proposed":[99],"evaluations":[100],"more":[101],"faithfully":[102],"reflect":[103],"perceived":[104],"generation":[105],"quality,":[106],"evidenced":[108],"by":[109],"stronger":[110],"correlations":[111],"human-rated":[113],"mean":[114],"opinion":[115],"scores":[116],"(MOS).":[117],"When":[118],"assessed":[119],"under":[120],"new":[122],"metrics,":[123],"relative":[125],"performance":[126],"landscape":[127],"is":[132,154],"reshaped,":[133],"revealing":[134],"significantly":[136],"reduced":[137],"gap":[138],"best-performing":[141],"model":[142],"human":[145],"topline.":[146],"Together,":[147],"results":[149],"suggest":[150],"critical":[155],"accurately":[157],"assessing":[158],"progress":[159],"modeling.":[163]},"counts_by_year":[],"updated_date":"2026-07-28T07:46:37.118299","created_date":"2026-01-14T00:00:00"}