Ivalin Chobanov

My research is in the area of reinforcement learning, with a focus on designing agents that can learn skills without extrinsic rewards that can generalize to open-ended environments. Particularly, I am interested in designing and utilizing latent world models and geometric constraints in order to ground the skill learning proceess in the dynamics, physics and physical properties of the environment, while respecting the properties of the desired skills. My further interests include instrinsic motivation, representation learning, generatve modelling and multi-agent systems.

I am a member of the Centre for AI in Assistive Autonomy and MARBLE, supervised by Subramanian Ramamoorthy and Ignacio Carlucho.

Before starting my PhD, I received a Master's degree in Artificial Intelligence from the University of Edinburgh and a Bachelor's degree in Mathematics with Computer Science from the University of Southampton. I have previously worked as a Software Developer at StoneX Financial Ltd. as part of the Metal's front-office team.