通过合成自然语言推理评估大语言模型中的多语言和语码转换对齐
Evaluating Multilingual and Code-Switched Alignment in LLMs via Synthetic Natural Language Inference
AI总结：
针对大语言模型跨语言逻辑对齐能力未充分探索的问题，提出合成逻辑型NLI对并翻译至多语言的受控评估框架，发现语码转换不降低甚至提升性能，为多语言鲁棒性研究提供新视角与工具。
摘要：
大语言模型（LLMs）在多语言场景中的应用日益广泛，但其跨语言保持一致、逻辑基础对齐的能力仍未得到充分探索。我们提出一种受控的多语言自然语言推理（NLI）评估框架，生成基于逻辑的合成前提-假设对，并将其翻译成类型多样的语言集合。该设计能精确控制语义关系，允许在单语和混合语言（语码转换）条件下测试。令人惊讶的是，语码转换不会降低甚至能提升性能，表明翻译引发的词汇变异可能充当正则化信号。我们通过基于嵌入的相似性分析和跨语言对齐可视化验证语义保留，确认翻译对的保真度。研究结果揭示当前LLM跨语言推理的潜力与脆弱性，指出语码转换是提升多语言鲁棒性的有前景手段。代码可在https://github.com/KurbanIntelligenceLab/nli-stress-testing获取。
英文摘要：
Large language models (LLMs) are increasingly applied in multilingual contexts, yet their capacity for consistent, logically grounded alignment across languages remains underexplored. We present a controlled evaluation framework for multilingual natural language inference (NLI) that generates synthetic, logic-based premise-hypothesis pairs and translates them into a typologically diverse set of languages. This design enables precise control over semantic relations and allows testing in both monolingual and mixed-language (code-switched) conditions. Surprisingly, code-switching does not degrade, and can even improve, performance, suggesting that translation-induced lexical variation may serve as a regularization signal. We validate semantic preservation through embedding-based similarity analyses and cross-lingual alignment visualizations, confirming the fidelity of translated pairs. Our findings expose both the potential and the brittleness of current LLM cross-lingual reasoning, and identify code-switching as a promising lever for improving multilingual robustness. Code available at: https://github.com/KurbanIntelligenceLab/nli-stress-testing
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。
