1 paper
Junlin Fang, Chong Zhang, Do Nguyen-Thanh +3
Large language model (LLM) agents must generalize from tools seen during training to unseen tools at deployment. A key challenge is tool trialing, i.e., excessive trials waste the…