Training on Artificially Code-Switched Data for Robust Multilingual LLMs in Cross-Lingual NLI

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

Research goal: How does training on artificially code-switched data affect the robustness of multilingual LLMs against adversarial perturbations in cross-lingual natural language inference tasks compared to monolingual training?

Autonomous synthesis report generated by Assignee Research. Tribunal consensus score: 7.5/10.