My research focuses on reinforcement learning from a theoretical perspective, particularly actor-critic and policy gradients. I enjoy finding connections between RL and other fields of machine learning; most recently, my research has explored the rich relationship between probabilistic inference and reinforcement learning.