Chinese Language

Research on the Chinese language is rapidly advancing, driven by the need to improve large language models' (LLMs) understanding and generation capabilities across diverse Chinese varieties, including Mandarin, Classical Chinese, and Taiwanese Hokkien. Current efforts focus on developing comprehensive benchmarks for evaluating LLMs on various tasks, from basic understanding to complex reasoning, and employing techniques like continual pre-training, instruction tuning, and knowledge grounding to enhance model performance. These advancements are crucial for bridging the resource gap in low-resource languages, improving cross-lingual understanding, and enabling new applications in education, cultural preservation, and other fields.