BengaliMoralBench：一种用于审计大型语言模型道德推理的基准，针对孟加拉语和文化
BengaliMoralBench: A Benchmark for Auditing Moral Reasoning in Large Language Models within Bengali Language and Culture
AI总结：
本文提出BengaliMoralBench，一个针对孟加拉语和文化背景的道德推理评估基准，涵盖五个道德领域，通过三种伦理框架进行标注，评估不同模型的性能差异及文化适应性问题。
摘要：
随着多语言大型语言模型（LLMs）在南亚的普及，其与本地伦理规范的对齐，尤其是孟加拉语，成为亟待探索的领域。现有伦理基准多以英语为中心，受西方道德框架影响，忽视了文化细微差别。为此，我们引入BengaliMoralBench，一个大规模伦理基准，专门针对孟加拉语和社会文化背景。该基准涵盖五个道德领域：日常活动、习惯、育儿、家庭关系和宗教活动，每个领域细分为十个文化相关的子类，总计50个子主题。每个场景通过母语者共识，通过三种伦理框架进行标注：德性伦理、常识伦理和正义伦理。我们通过统一提示协议，在开放权重和闭源模型上进行系统性零样本评估，包括最近的Llama和Gemma变体、Qwen和DeepSeek模型、前沿模型（GPT-4o-mini和Gemini 1.5 Pro）以及一个大型多语言基线（Qwen3-Next-80B）。结果表明，不同框架和领域间存在显著的性能差异，我们的定性分析揭示了文化基础、常识推理和道德公平性方面的持续弱点。这些发现暴露了当前LLMs在非西方环境中的关键限制，并突显了文化基础评估的必要性。BengaliMoralBench为负责任的本地化和基准测试提供了基础，以支持在文化多样性和低资源市场（如孟加拉国）部署语言技术。
英文摘要：
As multilingual Large Language Models (LLMs) gain traction across South Asia, their alignment with local ethical norms, particularly for Bengali, spoken by over 285 million people worldwide and among the most widely spoken languages globally, remains underexplored. Existing ethics benchmarks are predominantly English-centric and shaped by Western moral frameworks, overlooking cultural nuances vital for real-world deployment. To address this gap, we introduce BengaliMoralBench, a large-scale ethics benchmark designed for Bengali language and sociocultural contexts. Our benchmark spans five moral domains: (1) Daily Activities, (2) Habits, (3) Parenting, (4) Family Relationships, and (5) Religious Activities, each subdivided into ten culturally grounded categories, totaling 50 subtopics. Each scenario is annotated through native-speaker consensus under three ethical lenses: virtue ethics, commonsense ethics, and justice ethics. We conduct a systematic zero-shot evaluation under a unified prompting protocol across both open-weight and closed-source models, including recent Llama and Gemma variants, Qwen and DeepSeek models, frontier models (GPT-4o-mini and Gemini 1.5 Pro), and a large multilingual baseline (Qwen3-Next-80B). Results show substantial variation in performance across lenses and domains, and our qualitative analysis reveals persistent weaknesses in cultural grounding, commonsense reasoning, and moral fairness. These findings expose critical limitations of current LLMs in non-Western settings and underscore the need for culturally grounded evaluation. BengaliMoralBench provides a foundation for responsible localization and benchmarking to support the deployment of language technologies in culturally diverse, low-resource markets such as Bangladesh.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
