I am an Applied Research Scientist at Alibaba Group, where I work on LLM agent systems. My current research focuses on LLM post-training, agent memory, and reliable long-horizon task execution. Previously, at Huawei, I contributed to MindSpore, a deep learning framework; MindRLHF, a large-scale RLHF training framework; and the MetaERP agent, which streamlines invoice reconciliation and validation for accountants.

I received my M.Sc. in Computer Science and Technology from Tianjin University under Prof. Jianye Hao. My early research focused primarily on deep reinforcement learning, transfer learning, and multi-task reinforcement learning.

Outside of research, I enjoy photography, solving Rubik’s cubes, and traveling.