1 paper · 1 filter
Shiju Wang, Fei Ren, Fangcheng Fu +4
LLM agents execute long-horizon workflows where each model response determines the progress of subsequent tool interactions and environment transitions. Unlike chatbot serving, whe…