Memory, generalization, and the neural code.

The brain can store a lifetime of detailed memories while simultaneously creating abstract, generalized categories to quickly classify stimuli and experiences. How the brain implements these two conflicting processes to achieve flexible behavior and high memory capacity is still an open question.

In my research, I interrogate neural data using tools and ideas from machine learning and statistical physics to tackle this problem from a computational standpoint, investigating the geometrical coding properties of neural representations and their implications on cognition and memory.

As a physicist by training, I believe the finest scientific progress happens when experiments and theory guide each other. For this reason, my approach is strongly data-oriented, in closed-loop collaboration with experimental neuroscientists. Ongoing and recent collaborations include the Siegelbaum, Salzman, Kellendonk, and Hussaini Labs (Columbia), Schmidt-Hieber Lab (Institut Pasteur), Lopez-Roja Lab (UWM), and Rutishauser Lab (Caltech & Cedars-Sinai).

I received my Ph.D. in Statistical Physics from École Normale Superieure in Paris, under the supervision of Rémi Monasson and Simona Cocco. Currently, I am a postdoc in the group of Stefano Fusi in the Center for Theoretical Neuroscience of the Zuckerman Institute at Columbia University in New York.