1 paper
Chihiro Yano, Jun Hirako, Hiroaki Yamagiwa +1
We present J-tau, the first publicly available Japanese benchmark for evaluating tool-using LLM agents in interactive customer-support scenarios. While tau-bench is widely used to…