Benchmarking and Enhancing LLMs for Rule-Intensive Review of National Standard Documents

Tao Wang, Qihao Yang, Rongjiao Liang +4

Large language models (LLMs) increasingly support complex professional tasks, yet their capabilities in rule-intensive document review remain insufficiently evaluated. National sta…