About this role
Bring physics into world models - and make embodied agents fast, reliable and ready for the real world! World models are controllable, physics- and mechanism-grounded simulators of reality: they let an agent rehearse the consequences of its actions thousands of times before it acts in the real world. Paired with agents that plan, reason and act inside them, they are emerging as the substrate of Physical AI - and they cut across domains that look different on the surface but share the same scientific core: a robot rehearsing a manipulation task in a digital twin, and a biomedical model simulating how a tumour responds to therapy across biological scales, are both betting that structure beats brute scale. In this postdoc you will lead research on physics-informed (and physics-grounded) world models and agents: how to consolidate and transfer physical knowledge into learning systems for the real world. Physics engines and physics-informed neural networks are powerful, but the open problems are exactly where they stop being sufficient - under partial observation, in complex regimes that demand orders-of-magnitude speed-up, and where knowledge must transfer across embodiments, scenes and scientific domains. World models are controllable, physics- and mechanism-grounded simulators of reality: they let an agent rehearse the consequences of its actions thousands of times before it acts in the real world. Paired with agents that plan, reason and act inside them, they are emerging as the substrate of Physical AI - and they cut across domains that look different on the surface but share the same scientific core: a robot rehearsing a manipulation task in a digital twin, and a biomedical model simulating how a tumour responds to therapy across biological scales, are both betting that structure beats brute scale. In this postdoc you will lead research on physics-informed (and physics-grounded) world models and agents: how to consolidate and tr...