I work on robot learning, with a focus on reinforcement learning and imitation learning for robotics. I care about approaches that enable flexible physical intelligence in the real world, emphasizing robustness, reliability, and generalization. My research spans self-supervised and visual RL, RL fine-tuning of pretrained multimodal policies, and the training and evaluation of generalist robotics policies.

Beyond my research, I’m usually training for long-distance races, from 5Ks to marathons, and I also thoroughly enjoy playing and watching football. I’m particularly drawn to the adaptive motor skills of athletes, where coordination, reactivity, and improvisation come together. At the same time, I read widely across cognitive science, philosophy, scientific history, and fiction.