作者位次 #6/12ACL 2026Findings· Findings of ACL
Training Verifier to Assessing Complex Real-World Tool-Use Trajectories
Linzhuang Sun · Mingyang Chen · Hao Liang · Tianpeng Li · Zhou Yijie · Chenzheng Zhu · Tianyu Guo · Huanyao Zhang · Jingxuan Wei · Bihui Yu · Fan Yang · Wentao Zhang
摘要
Training effective AI agents for real-world tool-use interactions requires data that faithfully captures the dynamics of human–agent collaboration. However, such data is scarce, and existing methods often resort to synthetic data generation. The inherently dynamic and complex nature of user–agent...