{"id":"https://openalex.org/W7202286378","doi":"https://doi.org/10.48550/arxiv.2608.10414","title":"How Robust Are LLMs to Vietnamese Dialects?","display_name":"How Robust Are LLMs to Vietnamese Dialects?","publication_year":2026,"publication_date":"2026-08-11","ids":{"openalex":"https://openalex.org/W7202286378","doi":"https://doi.org/10.48550/arxiv.2608.10414"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2608.10414","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2608.10414","pdf_url":"https://arxiv.org/pdf/2608.10414","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,"listed_in":[],"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":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"volume":null,"type":"preprint","is_traditional":true,"indexed_in":["arxiv","datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/2608.10414","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5130389281","display_name":"Minh Tran","orcid":"https://orcid.org/0009-0008-7935-6881","observed_orcids":["https://orcid.org/0009-0008-7935-6881"]},"institutions":[{"id":"https://openalex.org/I23582244","display_name":"Ho Chi Minh City University of Science","ror":"https://ror.org/05jfbgm49","country_code":"VN","type":"education","lineage":["https://openalex.org/I123565023","https://openalex.org/I23582244"]}],"countries":["VN"],"is_corresponding":false,"raw_author_name":"Tran, Minh","raw_affiliation_strings":["University of Science , VNU-HCM ,"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"University of Science , VNU-HCM ,","institution_ids":["https://openalex.org/I23582244"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5130326928","display_name":"Trinh Chau","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I4405277059","display_name":"VNU University of Engineering and Technology","ror":"https://ror.org/019sx5d52","country_code":"VN","type":"education","lineage":["https://openalex.org/I177233841","https://openalex.org/I4405277059"]}],"countries":["VN"],"is_corresponding":false,"raw_author_name":"Chau, Trinh","raw_affiliation_strings":["VNU University of Engineering and Technology"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"VNU University of Engineering and Technology","institution_ids":["https://openalex.org/I4405277059"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5083870567","display_name":"Thanh-Nhan Le","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I120250893","display_name":"New York University Abu Dhabi","ror":"https://ror.org/00e5k0821","country_code":"AE","type":"education","lineage":["https://openalex.org/I120250893","https://openalex.org/I57206974"]}],"countries":["AE"],"is_corresponding":false,"raw_author_name":"Le, Thanh-Nhan","raw_affiliation_strings":["New York University Abu Dhabi ,"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"New York University Abu Dhabi ,","institution_ids":["https://openalex.org/I120250893"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5153546489","display_name":"Nam Tran","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Tran, Nam","raw_affiliation_strings":["Independent Researcher"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Independent Researcher","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5125646961","display_name":"Luan Thanh Nguyen","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I123565023","display_name":"Vietnam National University Ho Chi Minh City","ror":"https://ror.org/00waaqh38","country_code":"VN","type":"education","lineage":["https://openalex.org/I123565023"]},{"id":"https://openalex.org/I4387930302","display_name":"University of Information Technology","ror":"https://ror.org/05kdc4973","country_code":"VN","type":"education","lineage":["https://openalex.org/I123565023","https://openalex.org/I4387930302"]}],"countries":["VN"],"is_corresponding":false,"raw_author_name":"Nguyen, Luan Thanh","raw_affiliation_strings":["University of Information Technology , VNU-HCM"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"University of Information Technology , VNU-HCM","institution_ids":["https://openalex.org/I123565023","https://openalex.org/I4387930302"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5095057026","display_name":"Cuong Dang","orcid":"https://orcid.org/0009-0009-7463-0148","observed_orcids":["https://orcid.org/0009-0009-7463-0148"]},"institutions":[{"id":"https://openalex.org/I859038795","display_name":"Virginia Tech","ror":"https://ror.org/02smfhw86","country_code":"US","type":"education","lineage":["https://openalex.org/I859038795"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Dang, Cuong","raw_affiliation_strings":["Virginia Tech ,"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Virginia Tech ,","institution_ids":["https://openalex.org/I859038795"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5146990404","display_name":"Duc Hoang","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I63966007","display_name":"Massachusetts Institute of Technology","ror":"https://ror.org/042nb2s44","country_code":"US","type":"education","lineage":["https://openalex.org/I63966007"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Hoang, Duc","raw_affiliation_strings":["MIT"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"MIT","institution_ids":["https://openalex.org/I63966007"]}]}],"institutions":[],"countries_distinct_count":3,"institutions_distinct_count":7,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":true,"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":0.6093999743461609,"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":0.6093999743461609,"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/T10181","display_name":"Natural Language Processing Techniques","score":0.29269999265670776,"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/T12031","display_name":"Speech and dialogue systems","score":0.017100000753998756,"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/vietnamese-dialects","display_name":"Vietnamese dialects","score":0.9509999752044678},{"id":"https://openalex.org/keywords/dialect-variation","display_name":"dialect variation","score":0.6610000133514404},{"id":"https://openalex.org/keywords/emotion-recognition","display_name":"emotion recognition","score":0.5870000123977661},{"id":"https://openalex.org/keywords/natural-language-inference","display_name":"natural language inference","score":1.0},{"id":"https://openalex.org/keywords/question-answering","display_name":"question answering","score":0.996999979019165}],"concepts":[{"id":"https://openalex.org/C103621254","wikidata":"https://www.wikidata.org/wiki/Q9199","display_name":"Vietnamese","level":2,"score":0.968500018119812},{"id":"https://openalex.org/C63479239","wikidata":"https://www.wikidata.org/wiki/Q7353546","display_name":"Robustness (evolution)","level":3,"score":0.6557000279426575},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.5404999852180481},{"id":"https://openalex.org/C136886441","wikidata":"https://www.wikidata.org/wiki/Q926129","display_name":"Normalization (sociology)","level":2,"score":0.52920001745224},{"id":"https://openalex.org/C41895202","wikidata":"https://www.wikidata.org/wiki/Q8162","display_name":"Linguistics","level":1,"score":0.45339998602867126},{"id":"https://openalex.org/C185798385","wikidata":"https://www.wikidata.org/wiki/Q1161707","display_name":"Benchmark (surveying)","level":2,"score":0.4505999982357025},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.4345000088214874},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.4293999969959259},{"id":"https://openalex.org/C149782125","wikidata":"https://www.wikidata.org/wiki/Q160039","display_name":"Econometrics","level":1,"score":0.39989998936653137},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.3727000057697296},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.36890000104904175},{"id":"https://openalex.org/C2778334786","wikidata":"https://www.wikidata.org/wiki/Q1586270","display_name":"Variation (astronomy)","level":2,"score":0.367900013923645},{"id":"https://openalex.org/C27158222","wikidata":"https://www.wikidata.org/wiki/Q5532422","display_name":"Generalizability theory","level":2,"score":0.336899995803833},{"id":"https://openalex.org/C105795698","wikidata":"https://www.wikidata.org/wiki/Q12483","display_name":"Statistics","level":1,"score":0.3131999969482422},{"id":"https://openalex.org/C15744967","wikidata":"https://www.wikidata.org/wiki/Q9418","display_name":"Psychology","level":0,"score":0.31310001015663147},{"id":"https://openalex.org/C44291984","wikidata":"https://www.wikidata.org/wiki/Q1074173","display_name":"Question answering","level":2,"score":0.3018999993801117},{"id":"https://openalex.org/C195324797","wikidata":"https://www.wikidata.org/wiki/Q33742","display_name":"Natural language","level":2,"score":0.2815000116825104},{"id":"https://openalex.org/C2776230583","wikidata":"https://www.wikidata.org/wiki/Q1322198","display_name":"Spoken language","level":2,"score":0.25929999351501465},{"id":"https://openalex.org/C2987487971","wikidata":"https://www.wikidata.org/wiki/Q8096","display_name":"Lexical access","level":3,"score":0.2565999925136566},{"id":"https://openalex.org/C205649164","wikidata":"https://www.wikidata.org/wiki/Q1071","display_name":"Geography","level":0,"score":0.2502000033855438}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2608.10414","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2608.10414","pdf_url":"https://arxiv.org/pdf/2608.10414","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,"listed_in":[],"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":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"doi:10.48550/arxiv.2608.10414","doi":"https://doi.org/10.48550/arxiv.2608.10414","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2608.10414","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,"listed_in":[],"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"}],"best_oa_location":{"id":"pmh:oai:arXiv.org:2608.10414","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2608.10414","pdf_url":"https://arxiv.org/pdf/2608.10414","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,"listed_in":[],"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":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"sustainable_development_goals":[],"study_designs":[],"awards":[],"funders":[],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7202286378.pdf","grobid_xml":"https://content.openalex.org/works/W7202286378.grobid-xml"},"referenced_works_count":6,"referenced_works":["https://openalex.org/W4362515116","https://openalex.org/W4387561528","https://openalex.org/W4401306886","https://openalex.org/W4401306979","https://openalex.org/W4405655184","https://openalex.org/W6947991961"],"related_works":["https://openalex.org/W4416529938","https://openalex.org/W7140124191","https://openalex.org/W4416034762","https://openalex.org/W4403573135","https://openalex.org/W7140171337","https://openalex.org/W7168040497","https://openalex.org/W4416033566","https://openalex.org/W7137944660","https://openalex.org/W4416034421","https://openalex.org/W4416034603"],"abstract_inverted_index":{"Large":[0],"Language":[1],"Models":[2],"(LLMs)":[3],"are":[4,173],"typically":[5],"evaluated":[6,165],"on":[7,234],"standard":[8],"written":[9],"Vietnamese,":[10],"yet":[11],"everyday":[12],"communication":[13],"frequently":[14],"involves":[15],"regional":[16,244],"dialects":[17],"that":[18,132,231],"preserve":[19],"meaning":[20],"but":[21],"differ":[22],"in":[23,146],"surface":[24],"form.":[25],"Existing":[26],"Vietnamese":[27,45,62,92,99,236],"dialect":[28,63,93,187,213],"work":[29],"largely":[30],"addresses":[31],"this":[32,50],"issue":[33],"through":[34],"dialect-to-standard":[35],"normalization":[36],"instead":[37],"of":[38,58],"measuring":[39],"how":[40],"the":[41,54,133,178,193,218],"model":[42,86,130,166],"fails":[43],"under":[44,242],"dialectal":[46,105,134,156],"inputs.":[47],"To":[48],"address":[49],"gap,":[51],"we":[52],"present":[53],"first":[55],"systematic":[56],"evaluation":[57,124],"LLM":[59],"robustness":[60],"to":[61,148],"variation":[64],"across":[65,90,186,223],"multiple":[66],"tasks,":[67],"quantifying":[68],"performance":[69,160,196,208,233],"degradation":[70],"and":[71,102,118,163,190,200],"failure":[72],"patterns.":[73],"We":[74],"introduce":[75],"VialectBench":[76,95],"(Vietnamese":[77],"Dialects":[78],"Benchmarking),":[79],"a":[80,126,137],"controlled":[81],"benchmark":[82],"for":[83],"testing":[84],"whether":[85],"decisions":[87],"remain":[88],"stable":[89],"six":[91],"groups.":[94],"contains":[96],"400":[97],"Standard":[98,150,235],"source":[100],"instances":[101],"2,400":[103],"human-written":[104],"rewrites":[106,135],"spanning":[107],"emotion":[108],"recognition":[109],"(ER),":[110],"natural":[111],"language":[112,129],"inference":[113],"(NLI),":[114],"question":[115,120],"answering":[116,121],"(QA),":[117],"multiple-choice":[119],"(MCQA).":[122],"Dataset":[123],"with":[125,175],"fixed":[127],"reference":[128],"shows":[131],"induce":[136],"measurable":[138],"model-relative":[139],"likelihood":[140],"shift":[141],"while":[142],"remaining":[143],"nearly":[144],"equal":[145],"length":[147],"their":[149],"counterparts.":[151],"Across":[152],"ten":[153],"instruction-tuned":[154],"models,":[155,225],"inputs":[157],"reduce":[158],"average":[159,180,195,207,220],"by":[161,209],"2.82%,":[162],"no":[164],"is":[167],"fully":[168],"dialect-invariant.":[169],"All":[170],"four":[171],"tasks":[172],"affected,":[174],"QA":[176],"showing":[177],"largest":[179,194],"degradation.":[181],"Robustness":[182],"also":[183,216],"varies":[184],"substantially":[185],"groups:":[188],"PNT3":[189],"PNT2":[191],"cause":[192],"drops,":[197],"at":[198,226],"6.17%":[199],"4.73%,":[201],"respectively,":[202],"whereas":[203],"PNB":[204],"slightly":[205],"improves":[206],"0.42%.":[210],"The":[211],"Central":[212],"group":[214],"(PNT1-PNT4)":[215],"yields":[217],"highest":[219],"harmful-flip":[221],"rate":[222],"all":[224],"6.54%.":[227],"These":[228],"findings":[229],"show":[230],"strong":[232],"does":[237],"not":[238],"guarantee":[239],"reliable":[240],"behavior":[241],"meaning-preserving":[243],"variation.":[245]},"counts_by_year":[],"updated_date":"2026-11-07T11:25:07.637180","created_date":"2026-08-13T00:00:00"}
