Philip Ball

Latest

• Learning General World Models in a Handful of Reward-Free Deployments
• Towards Tractable Optimism in Model-Based Reinforcement Learning
• Ready Policy one: World Building Through Active Learning