关于

清华大学计算机系 KEG 实验室博一在读，研究方向为大模型强化学习（LLM Reinforcement Learning）、深度研究智能体（Deep Search Agent）、多智能体（Multi-Agent）。本科毕业于清华大学未央书院（理工双学位，数理基础科学+软件工程），欢迎添加微信（微信号 learning_rate）和我交流～😊

研究成果：

• GLM-5（核心贡献者，GitHub ~2k Stars）

• DeepDive（第一作者，GitHub ~300 Stars，Huggingface ~10k Downloads 📥）

• GLM-4.5/4.6/4.7（核心贡献者，GitHub ~4k Stars）

• TreeRL（共同一作，ACL 2025 Main）

• AgentTuning（共同一作，ACL 2024 Findings，GitHub ~1.5k Stars，Huggingface ~20k Downloads 📥）

---

I’m a first-year PhD student at the KEG Lab, Department of Computer Science, Tsinghua University. My research focuses on reinforcement learning for large language models, deep agent research and multi-agent. I completed my undergraduate studies at Weiyang College, Tsinghua University, earning dual bachelor’s degrees — a B.Sc. in Mathematics and Physics and a B.Eng. in Software Engineering. Feel free to connect with me on WeChat (ID: learning_rate) — happy to chat! 😊

Publications:

• GLM-5 (Core Contributor, GitHub ~2k Stars)

• DeepDive (First Author, GitHub ~300 Stars, Huggingface ~10k Downloads)

• GLM-4.5 (Core Contributor, GitHub ~4k Stars)

• TreeRL (Co-First Author, ACL 2025 Main)

• AgentTuning (Co-First author, ACL 2024 Findings, GitHub ~1.5k Stars, Huggingface ~20k Downloads 📥）