Yoonsang Lee

Hello! I am a Ph.D. student in Computer Science at Princeton University, advised by Danqi Chen. Prior to that, I did my undergrad at Seoul National University, fortunate to be working with Eunsol Choi.

My research focuses on LLMs and agents. Recently, I've been working on:

• Long-horizon agents: building agents that operate over hundreds of steps through better tool use, long-context reasoning, and domain-specific training (SLIM).
• Recursive self-improvement: scaling agentic systems at test time and enabling them to learn from their own trajectories and dense feedback (AggAgent).
• Retrieval-augmented LMs: developing models that seek real-world information beyond their training data (AmbigDocs, RARe, EGG).

Feel free to reach out for collaboration!