Large language model (LLM) agents are increasingly proposed for enterprise workflows, yet existing evaluations rarely test whether business-decision conclusi…
机构:上交
来源:arXiv 2609.04667 | AI4Papers 论文推荐平台