Joe Watson

Joe Watson is a Postdoctoral Researcher at the University of Oxford within the Applied Artificial Intelligence Lab at the Oxford Robotics Institute. He earned his PhD in 2024 from the Technical University of Darmstadt, supervised by Prof. Jan Peters, and holds an MEng in Information & Computer Engineering from University of Cambridge.

His research focuses on statistical methods for robot learning, particularly the duality between entropy-regularized optimization and Bayesian inference. Key interests include uncertainty quantification, policy optimization, and incorporating structural knowledge into robotics.

Scientific contributions include Coherent Soft Imitation Learning (NeurIPS 2023, spotlight) and Monte Carlo Posterior Policy Iteration (CoRL 2022, oral). Recent publications address safe foundation models and tensor motion planning in 2025.

• 2022: R:SS Pioneer for robotics research
• 2016: Charles Babbage Senior Scholarship at Cambridge

He has contributed to open-source projects like i2c and monte-carlo-posterior-policy-iteration, with 534 GitHub contributions in the last year.