How Robust Are LLMs to Vietnamese Dialects?

• URL: http://arxiv.org/abs/2608.10414v1
• Date: Tue, 11 Aug 2026 03:02:57 GMT
• Title: How Robust Are LLMs to Vietnamese Dialects?
• Abstract summary: We present the first systematic evaluation of Large Language Models (LLMs) evaluating Vietnamese to Vietnamese dialect variation across multiple tasks.<n>VialectBench contains 400 Standard Vietnamese source instances and 2,400 human-written dialectal rewrites spanning emotion recognition (ER), natural language inference (NLI), question answering (QA), and multiple-choice question answering (MCQA)<n>Across ten instruction-tuned models, dialectal inputs reduce average performance by 2.82%, and no evaluated model is fully dialect-invariant.
• Score: 7.401730363695013
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Large Language Models (LLMs) are typically evaluated on standard written Vietnamese, yet everyday communication frequently involves regional dialects that preserve meaning but differ in surface form. Existing Vietnamese dialect work largely addresses this issue through dialect-to-standard normalization instead of measuring how the model fails under Vietnamese dialectal inputs. To address this gap, we present the first systematic evaluation of LLM robustness to Vietnamese dialect variation across multiple tasks, quantifying performance degradation and failure patterns. We introduce VialectBench (Vietnamese Dialects Benchmarking), a controlled benchmark for testing whether model decisions remain stable across six Vietnamese dialect groups. VialectBench contains 400 Standard Vietnamese source instances and 2,400 human-written dialectal rewrites spanning emotion recognition (ER), natural language inference (NLI), question answering (QA), and multiple-choice question answering (MCQA). Dataset evaluation with a fixed reference language model shows that the dialectal rewrites induce a measurable model-relative likelihood shift while remaining nearly equal in length to their Standard counterparts. Across ten instruction-tuned models, dialectal inputs reduce average performance by 2.82%, and no evaluated model is fully dialect-invariant. All four tasks are affected, with QA showing the largest average degradation. Robustness also varies substantially across dialect groups: PNT3 and PNT2 cause the largest average performance drops, at 6.17% and 4.73%, respectively, whereas PNB slightly improves average performance by 0.42%. The Central dialect group (PNT1-PNT4) also yields the highest average harmful-flip rate across all models, at 6.54%. These findings show that strong performance on Standard Vietnamese does not guarantee reliable behavior under meaning-preserving regional variation.

Related papers

• IndicQE-APE: A Benchmark for Quality Estimation and Automatic Post-Editing for Indic Languages [77.88427845181154]
  We consolidate the WMT 2020--2024 shared-task lineage with an extended English--Malayalam resource into indicqe: $126,754$ instances over nine directional pairs.<n>We benchmark six prompted LLMs and three COMET metrics on segment-level QE, and three systems on APE.
  arXiv  Detail & Related papers  (2026-08-17T09:50:52Z)
• DIA-HARM: Dialectal Disparities in Harmful Content Detection Across 50 English Dialects [6.107850985025956]
  We present DIA-HARM, the first benchmark for evaluating disinformation detection across 50 English dialects.<n>Human-written dialectal content degrades detection by 1.4-3.6% F1, while AI-generated content remains stable.<n>We release the DIA-HARM framework, D3 corpus, and evaluation tools.
  arXiv  Detail & Related papers  (2026-04-07T01:43:48Z)
• Benchmarking Bengali Dialectal Bias: A Multi-Stage Framework Integrating RAG-Based Translation and Human-Augmented RLAIF [0.3227658251731014]
  Large language models (LLMs) frequently exhibit performance biases against regional dialects of low-resource languages.<n>We propose a two-phase framework to evaluate dialectal bias in LLM question-answering across nine Bengali dialects.
  arXiv  Detail & Related papers  (2026-03-22T18:44:57Z)
• ViDia2Std: A Parallel Corpus and Methods for Low-Resource Vietnamese Dialect-to-Standard Translation [5.3220011447194215]
  We introduce ViDia2Std, the first manually annotated parallel corpus for dialect-to-standard Vietnamese translation.<n>Unlike prior datasets, ViDia2Std includes diverse dialects from Central, Southern, and non-standard Northern regions.<n>We report agreement rates of 86% (North), 82% (Central), and 85% (South) for ViDia2Std.
  arXiv  Detail & Related papers  (2026-03-10T20:24:38Z)
• DialectGen: Benchmarking and Improving Dialect Robustness in Multimodal Generation [111.94720088481614]
  Can multimodal generative models effectively produce content given dialectal textual input?<n>We construct a new large-scale benchmark spanning six common English dialects.<n>We design a general encoder-based mitigation strategy for multimodal generative models.
  arXiv  Detail & Related papers  (2025-10-16T17:56:55Z)
• Assessing Dialect Fairness and Robustness of Large Language Models in Reasoning Tasks [68.33068005789116]
  We introduce ReDial, a benchmark containing 1.2K+ parallel query pairs in Standardized English and AAVE.<n>We evaluate widely used models, including GPT, Claude, Llama, Mistral, and the Phi model families.<n>Our work establishes a systematic and objective framework for analyzing LLM bias in dialectal queries.
  arXiv  Detail & Related papers  (2024-10-14T18:44:23Z)
• Modeling Orthographic Variation in Occitan's Dialects [3.038642416291856]
  Large multilingual models minimize the need for spelling normalization during pre-processing. Our findings suggest that large multilingual models minimize the need for spelling normalization during pre-processing.
  arXiv  Detail & Related papers  (2024-04-30T07:33:51Z)
• Quantifying the Dialect Gap and its Correlates Across Languages [69.18461982439031]
  This work will lay the foundation for furthering the field of dialectal NLP by laying out evident disparities and identifying possible pathways for addressing them through mindful data collection.
  arXiv  Detail & Related papers  (2023-10-23T17:42:01Z)
• Unlikelihood Tuning on Negative Samples Amazingly Improves Zero-Shot Translation [79.96416609433724]
  Zero-shot translation (ZST) aims to translate between unseen language pairs in training data. The common practice to guide the zero-shot language mapping during inference is to deliberately insert the source and target language IDs. Recent studies have shown that language IDs sometimes fail to navigate the ZST task, making them suffer from the off-target problem.
  arXiv  Detail & Related papers  (2023-09-28T17:02:36Z)
• Multi-VALUE: A Framework for Cross-Dialectal English NLP [49.55176102659081]
  Multi- Dialect is a controllable rule-based translation system spanning 50 English dialects. Stress tests reveal significant performance disparities for leading models on non-standard dialects. We partner with native speakers of Chicano and Indian English to release new gold-standard variants of the popular CoQA task.
  arXiv  Detail & Related papers  (2022-12-15T18:17:01Z)
• Understanding by Understanding Not: Modeling Negation in Language Models [81.21351681735973]
  Negation is a core construction in natural language. We propose to augment the language modeling objective with an unlikelihood objective that is based on negated generic sentences. We reduce the mean top1 error rate to 4% on the negated LAMA dataset.
  arXiv  Detail & Related papers  (2021-05-07T21:58:35Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.