Building LLM agents that can plan, reason, and improve through reinforcement learning.

Ph.D. student at Tsinghua University · Researcher / Builder in Agentic RL and LLM post-training

I work on language model agents that need to operate beyond static benchmarks, with a focus on planning, reasoning, and learning under real-world system constraints.

My current interests include:

• Agentic reinforcement learning
• LLM post-training systems
• Long-horizon reasoning and planning
• Efficient infrastructure for open models

• Ph.D. student in Computer Science at Tsinghua University
• Advised by Prof. Hongning Wang
• Intern at Zhipu AI

Previously worked with Shanghai AI Lab, HKU, and KAUST.

• Core contributor to GLM-5
• Core contributor to GLM-4.5
• Core contributor to slime, an LLM post-training framework for scalable RL
• Contributor to SGLang, mainly on RL-related capabilities

• GLM-5: from vibe coding to agentic engineering
• GLM-4.5: agentic, reasoning, and coding foundation models
• SWE-Fixer
  • ACL 2025 Findings
• Can Large Language Model Agents Simulate Human Trust Behavior?
  • NeurIPS 2024

I build practical LLM agent systems:
better reasoning, better training, better deployment.

• Homepage: https://yitianlian.github.io/
• GitHub: https://github.com/yitianlian