1 paper · 2 filters
Chengyuan Liu, Shihang Wang, Lizhi Qing +4
Domain Large Language Models (LLMs) are developed for domain-specific tasks based on general LLMs. But it still requires professional knowledge to facilitate the expertise for some…