I recently joined Google DeepMind Zürich as a Research Scientist and am wrapping up my PhD in the International Max Planck Research School for Intelligent Systems (IMPRS-IS).

My current research investigates how generative video models can improve multi-modal reasoning and visual intelligence.

The underlying theme of my PhD was how to learn robust representations—a topic I epxlored theoratically and empirically in large vision-language, language, and video models. Throughout this time, I was fortunate to work with and be guided by Wieland Brendel, Matthias Bethge, as well as Robert Geirhos and Priyank Jaini during my 2025 internship at Google DeepMind Toronto.