1 paper · 1 filter
Pengyu Dai, Weihao Xuan, Junjue Wang +4
Recent advances have enabled large language model (LLM) agents to solve complex tasks by orchestrating external tools. However, these agents often struggle in specialized, tool-int…