Culturally-Adapted Red-Teaming Across East and Southeast Asian Contexts: A Methodological and Comparative Analysis

• URL: http://arxiv.org/abs/2606.09178v1
• Date: Mon, 08 Jun 2026 08:17:18 GMT
• Title: Culturally-Adapted Red-Teaming Across East and Southeast Asian Contexts: A Methodological and Comparative Analysis
• Abstract summary: We construct paired DT and culturally-adapted datasets via 1:1 seed matching for four languages.<n>We compare Attack Success Rate (ASR) and Cultural Realism scores across four open-source safety evaluations.
• Score: 2.8523456804049885
• License: http://creativecommons.org/licenses/by-nc-sa/4.0/
• Abstract: Multilingual safety evaluation of large language models (LLMs) has predominantly relied on direct translation (DT) of English benchmarks into target languages - an approach that converts surface-level linguistic form while failing to reflect the cultural context embedded in threat scenarios, social norms, and legal frameworks. We construct paired DT and culturally-adapted (CA) datasets via 1:1 seed matching for four languages - Korean (KO), Japanese (JA), Thai (TH), and Khmer (KM) - and compare Attack Success Rate (ASR) and Cultural Realism scores across four open-source LLM. CA prompts yield Delta-ASR > 0 across all 16 language x model combinations (mean +9.3 pp), and DT-based evaluation underestimates risk in 44 of 48 category x language combinations. Language-level analysis reveals that the distribution of threat forms is heterogeneous across languages. Cultural Realism analysis further shows that DT Cultural Depth (C3) scores remain consistently below 1.0 out of 3.0 across all four languages (mean 0.17), whereas CA scores reach up to 2.51, indicating that direct translation produces inputs systematically divergent from those encountered in real-world multicultural settings. These findings demonstrate that adapting benchmarks to language-specific cultural contexts - rather than relying on linguistic translation alone - is necessary for valid multilingual LLM safety evaluation.

Related papers

• On the Limitations of Cross-Lingual Consistency in Multilingual Text-to-image Generation [79.18001266925245]
  LingT2I is a benchmark covering 10 widely used languages with 33K prompts.<n>We conduct a comprehensive cross-lingual analysis, uncovering linguistic inequality and language-dependent trade-offs.
  arXiv  Detail & Related papers  (2026-08-11T14:49:23Z)
• ROK-FORTRESS: Measuring the Effect of Geopolitical Transcreation for National Security and Public Safety [2.929429337200341]
  Safety evaluations for large language models increasingly target high-stakes National Security and Public Safety (NSPS) risks.<n> empirical evidence of how language and geopolitical context interact remains limited to a narrow set of language pairs.<n>We introduce emphROK-FORTRESS, a bilingual, culturally adversarial NSPS benchmark that uses the English--Korean language pair and U.S.--ROK geopolitical axis as a case study.
  arXiv  Detail & Related papers  (2026-05-13T22:07:22Z)
• Multilingual Refusal Alignment for Safer Large Language Models [53.64286756804503]
  We investigate whether single-language alignment transfers cross-lingually, how language consistency is preserved during training, and the resulting trade-offs with general knowledge capabilities.<n>We introduce RefusEU, a novel refusal alignment dataset covering 12 European languages, including a dedicated test set for evaluating current state-of-the-art models.<n>Our controlled Direct Preference Optimization (DPO) experiments provide two key insights: aligning models exclusively in English is insufficient to ensure cross-lingual safety, even for the same harm categories, whereas training on multilingual datasets can improve safety without degrading general performance, as measured by the Global MMLU benchmark.
  arXiv  Detail & Related papers  (2026-04-24T09:29:14Z)
• The GaoYao Benchmark: A Comprehensive Framework for Evaluating Multilingual and Multicultural Abilities of Large Language Models [51.61416200800499]
  GaoYao is a comprehensive benchmark with 182.3k samples, 26 languages and 51 nations/areas.<n>First, GaoYao proposes a unified framework categorizing evaluation tasks into three cultural layers.<n>Second, we achieve native-quality expansion by leveraging experts to rigorously localize subjective benchmarks into 19 languages.<n>Third, we conduct an in-depth diagnostic analysis on 20+ flagship and compact LLMs.
  arXiv  Detail & Related papers  (2026-04-22T06:19:46Z)
• LASA: Language-Agnostic Semantic Alignment at the Semantic Bottleneck for LLM Safety [88.98698230989186]
  Large language models (LLMs) often demonstrate strong safety performance in high-resource languages, yet exhibit severe vulnerabilities in low-resource languages.<n>We attribute this gap to a mismatch between language-agnostic semantic understanding ability and language-dominant safety alignment biased toward high-resource languages.<n>We propose Language-Agnostic Semantic Alignment (LASA), which anchors safety alignment directly in semantic bottlenecks.
  arXiv  Detail & Related papers  (2026-04-13T15:59:50Z)
• "Be My Cheese?": Cultural Nuance Benchmarking for Machine Translation in Multilingual LLMs [0.0]
  We present a large-scale human evaluation benchmark for assessing cultural localisation in machine translation.<n>We evaluate 7 multilingual large language models (LLMs) across 15 target languages with 5 native-speaker raters per language.<n>Across full-text evaluations, mean overall quality is modest (1.68/3): GPT-5 (2.10/3), Claude Sonnet 3.7 (1.97/3), and Mistral Medium 3.1 (1.84/3) form the strongest tier with fewer catastrophic failures.
  arXiv  Detail & Related papers  (2026-02-04T16:35:48Z)
• Improving Methodologies for LLM Evaluations Across Global Languages [19.63570354411416]
  The exercise shows how safety behaviours can vary across languages.<n>It also generated insights for improving multilingual safety evaluations.<n>This work represents an initial step toward a shared framework for multilingual safety testing of advanced AI systems.
  arXiv  Detail & Related papers  (2026-01-22T07:18:08Z)
• Evaluating Modern Large Language Models on Low-Resource and Morphologically Rich Languages:A Cross-Lingual Benchmark Across Cantonese, Japanese, and Turkish [12.286855282078305]
  GPT-4o, GPT-4, Claude3.5Sonnet, LLaMA3.1, MistralLarge2, LLaMA-2Chat13B, and Mistral7BInstruct are evaluated.<n>Our benchmark spans four diverse tasks: open-domain question answering, document summarization, English-to-X translation, and culturally grounded dialogue.
  arXiv  Detail & Related papers  (2025-11-05T22:09:53Z)
• CRaFT: An Explanation-Based Framework for Evaluating Cultural Reasoning in Multilingual Language Models [0.42970700836450487]
  We introduce CRaFT, an explanation-based multilingual evaluation framework designed to assess how large language models (LLMs) reason across cultural contexts.<n>We apply the framework to 50 culturally grounded questions from the World Values Survey, translated into Arabic, Bengali, and Spanish, and evaluate three models (GPT, DeepSeek, and FANAR) across over 2,100 answer-explanation pairs.<n>Results reveal significant cross-lingual variation in reasoning: Arabic reduces fluency, Bengali enhances it, and Spanish remains largely stable.
  arXiv  Detail & Related papers  (2025-10-15T18:49:10Z)
• I Am Aligned, But With Whom? MENA Values Benchmark for Evaluating Cultural Alignment and Multilingual Bias in LLMs [5.060243371992739]
  We introduce MENAValues, a novel benchmark designed to evaluate the cultural alignment and multilingual biases of large language models (LLMs)<n> Drawing from large-scale, authoritative human surveys, we curate a structured dataset that captures the sociocultural landscape of MENA with population-level response distributions from 16 countries.<n>Our analysis reveals three critical phenomena: "Cross-Lingual Value Shifts" where identical questions yield drastically different responses based on language, "Reasoning-Induced Degradation" where prompting models to explain their reasoning worsens cultural alignment, and "Logit Leakage" where models refuse sensitive questions while internal probabilities reveal strong hidden
  arXiv  Detail & Related papers  (2025-10-15T05:10:57Z)
• MMA-ASIA: A Multilingual and Multimodal Alignment Framework for Culturally-Grounded Evaluation [91.22008265721952]
  MMA-ASIA centers on a human-curated, multilingual, and multimodally aligned benchmark covering 8 Asian countries and 10 languages.<n>This is the first dataset aligned at the input level across three modalities: text, image (visual question answering), and speech.<n>We propose a five-dimensional evaluation protocol that measures: (i) cultural-awareness disparities across countries, (ii) cross-lingual consistency, (iii) cross-modal consistency, (iv) cultural knowledge generalization, and (v) grounding validity.
  arXiv  Detail & Related papers  (2025-10-07T14:12:12Z)
• Code-Switching In-Context Learning for Cross-Lingual Transfer of Large Language Models [64.54005959758733]
  We introduce code-switching in-context learning (CSICL) as a principled and robust approach for overcoming the translation barrier during inference.<n>We conduct extensive experiments across 4 LLMs, 6 datasets, and 10 languages, spanning both knowledge-intensive and reasoning-oriented domains.<n>Our results demonstrate CSICL consistently outperforms X-ICL baselines, achieving gains of 3.1%p and 1.9%p in both target and unseen languages.
  arXiv  Detail & Related papers  (2025-10-07T08:35:42Z)
• XIFBench: Evaluating Large Language Models on Multilingual Instruction Following [59.549015333755186]
  Large Language Models (LLMs) have demonstrated remarkable instruction-following capabilities across various applications.<n>Existing evaluations lack fine-grained constraint analysis across diverse linguistic contexts.<n>We introduce XIFBench, a comprehensive benchmark for evaluating multilingual instruction-following abilities of LLMs.
  arXiv  Detail & Related papers  (2025-03-10T17:07:52Z)
• LLMs Lost in Translation: M-ALERT uncovers Cross-Linguistic Safety Inconsistencies [63.10843814055688]
  M-ALERT is a benchmark that evaluates the safety of Large Language Models in five languages.<n>M-ALERT includes 15k high-quality prompts per language, totaling 75k, with category-wise annotations.<n>Our experiments on 39 state-of-the-art LLMs highlight the importance of language-specific safety analysis.
  arXiv  Detail & Related papers  (2024-12-19T16:46:54Z)

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.