• Ph.D. Student in Machine Learning and Control at Hybrid Robotics Lab, BAIR, UC Berkeley
• Working on Reinforcement Learning and Stochastic Control

Safe and Robust RL, Continual RL, Stochastic Control, Data-Efficient RL, Optimization

• How can agents achieve safety, robustness, stability, sample-efficiency, and adaptation under non-stationary dynamics?
• Can agents performance a.s. monotone increase using any data stream? (= Is RL scalable?)
• Can agents create and/or discover alpha in the market?

Recoverable RL, World Models, Offline RL, Off2On RL, Off-Policy Q-Learning, Stochastic Control for Finance, Deep Learning for Dynamical Systems