I am a postdoctoral researcher with Eric Schulz at Helmholtz Munich and Peter Dayan at the Max Planck Institute for Biological Cybernetics. Before moving to Germany, I completed my Ph.D. in Computer and Communication Sciences at EPFL, under the supervision of Wulfram Gerstner. 

My research interests lie at the interface of Cognitive Neuroscience and Machine Learning. I am interested in questions like "What computational problems do we solve on a daily basis?", "What strategies do we use to solve these problems?", and "What can we learn from our behavior to develop the next generation of robust and safe AI?".

In recent years, I have explored how we employ signals related to surprise and novelty when facing complex computational problems. I have studied the mathematical characteristics of these signals and have identified how they guide learning and curiosity. You can find an accessible summary of some of our key results in this article by EPFL. Most recently, in my postdoctoral research, I have expanded my work to investigate the notions of controllability, agency, and empowerment. I have been exploring how to quantify the controllability of an environment and how an environment's controllability determines our subjective sense of control (e.g., see our preprint on PsyArXiv).