XIFBench: Evaluating Large Language Models on Multilingual Instruction Following
Li, Zhenyu, Chen, Kehai, Long, Yunfei, Bai, Xuefeng, Zhang, Yaoyin, Wei, Xuchen, Li, Juntao, Zhang, Min
–arXiv.org Artificial Intelligence
Large Language Models (LLMs) have demonstrated remarkable instruction-following capabilities across various applications. However, their performance in multilingual settings remains poorly understood, as existing evaluations lack fine-grained constraint analysis. We introduce XIFBench, a comprehensive constraint-based benchmark for assessing multilingual instruction-following abilities of LLMs, featuring a novel taxonomy of five constraint categories and 465 parallel instructions across six languages spanning different resource levels. To ensure consistent cross-lingual evaluation, we develop a requirement-based protocol that leverages English requirements as semantic anchors. These requirements are then used to validate the translations across languages. Extensive experiments with various LLMs reveal notable variations in instruction-following performance across resource levels, identifying key influencing factors such as constraint categories, instruction complexity, and cultural specificity.
arXiv.org Artificial Intelligence
Mar-10-2025
- Country:
- Asia > China (0.28)
- Europe > Middle East
- Malta (0.14)
- North America > United States (0.46)
- Genre:
- Research Report > New Finding (0.46)
- Industry:
- Education > Curriculum (0.47)
- Technology: