From a single synapse to simple behaviors, the nervous system exhibits rich and complex dynamics that convey every perception we experience and every action we make. I’m interested in understanding the basic mechanisms that underlie our sensations of the world with a particular interest in the tactile and visual senses. I was originally trained as a theoretical physicist and after a master thesis at the EPFL (Lausanne Switzerland), I decided to begin my PhD work in neuroscience. During my 4 years in Alain Destexhe lab at UNIC-CNRS (Gif-sur-Yvette, France), I had the opportunity to collaborate with many talented scientists who helped me develop my skills in computational and experimental neurosciences. Understanding how the brain behaves in absence of external stimulations and how this spontaneous activity reflect internal states, organize our memory and thoughts and interact with the sensory input coming from the external world is a key step for capturing the building blocks of the mind. My work was mainly focused in characterizing this spontaneous activity in vivo, observed during anesthetized states, sleep or quiet wakefulness. Using modeling approaches as well as advanced data analysis of electrophysiological recordings, I studied how coding and learning can be performed in these apparently stochastic network states. In particular, I studied how simple plasticity rules acting at the synaptic level could stabilize new memories within large neuron networks without being erased by the ongoing activity or without evoking pathological regimes.
These projects in collaboration with Yves Frégnac lab made me realize the importance of working in interaction with experimentalists and the need of developing a common language that could ensure the success of projects combining both cutting-edge experimental and computational approaches. Near the end of my graduate work, I became increasingly interested in learning experimental techniques in order to be able to design my own experiments and gather data that would be consistent with model-based analysis. I started a collaboration with Daniel Shulz lab where I was trained to perform in vivo surgeries as well as multielectrode recordings in the rat somatosensory cortex. Using reverse-correlation techniques, we showed how functionally distinct neuron subclasses in the barrel cortex encode the properties of tactile scenes.