{"ID":2880282,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2508.14735","arxiv_id":"2508.14735","title":"Evaluating Multilingual and Code-Switched Alignment in LLMs via Synthetic Natural Language Inference","abstract":"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","short_abstract":"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 pre...","url_abs":"https://arxiv.org/abs/2508.14735","url_pdf":"https://arxiv.org/pdf/2508.14735v1","authors":"[\"Samir Abdaljalil\",\"Erchin Serpedin\",\"Khalid Qaraqe\",\"Hasan Kurban\"]","published":"2025-08-20T14:30:34Z","proceeding":"cs.CL","tasks":"[\"cs.CL\",\"cs.AI\"]","methods":"[\"Large Language Model\",\"Language Model\"]","has_code":false,"code_links":[{"ID":610665,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_id":2880282,"paper_url":"https://arxiv.org/abs/2508.14735","paper_title":"Evaluating Multilingual and Code-Switched Alignment in LLMs via Synthetic Natural Language Inference","repo_url":"https://github.com/KurbanIntelligenceLab/nli-stress-testing","is_official":false,"mentioned_in_paper":false,"mentioned_in_github":true,"github_stars":0}]}