Wen Sun

Wen Sun is an Assistant Professor in the Computer Science Department at Cornell University and a research scientist at Databricks. He holds an adjunct role at Cornell Tech. Previously, he was a postdoctoral researcher at Microsoft Research NYC (2019–2020) and earned his PhD from Carnegie Mellon University's Robotics Institute under Drew Bagnell's supervision. His research focuses on Reinforcement Learning (RL), AI, and decision-making, with recent work emphasizing RL for generative models, representation learning, and theoretical foundations of RLHF (Reinforcement Learning from Human Feedback). His group develops software tools like the RL via generative models library and explores hybrid approaches combining offline and online data.

Education:

• PhD in Robotics, Carnegie Mellon University (2019)
• MS and BS in Computer Science (implied from career progression)

Research Interests:

Dr. Sun's work spans core RL theory, applications in generative models, and algorithmic innovation. Key areas include:

• RL for fine-tuning LLMs/diffusion models
• Representation learning in RL
• Efficient policy optimization
• Risk-sensitive decision-making
• Pac RL in POMDPs and low-rank MDPs

Teaching:

• CS 4789/5789: Intro to Reinforcement Learning (Spring 2025, 2021)
• CS 6789: Foundations of Reinforcement Learning (Fall 2024)
• CS 4780/5780: Intro to Machine Learning (Fall 2023)

Key Awards:

• Sloan Research Fellowship (2025)
• NSF Career Award (2024)
• Ann S. Bowers Research Excellence Award (2024)

Grants & Labs:

His research is supported by NSF and industry collaborations. The lab maintains active software projects, including tools for RL in generative models and database query optimization (JoinGym). Collaborations involve institutions like UW, UIUC, and Databricks.

Future Work:

Expanding RL applications in LLM reasoning, improving generative model efficiency, and advancing theoretical guarantees for hybrid RLHF frameworks.