Computational Models of Neural Representations and Computations
Alex Pouget
Our research focuses on two main topics: neural coding and spatial representations. The goal of the work on neural coding is to understand how neurons encode information, such as the color of an object or the direction of the next hand movement, and how computation is carried out in the cortical circuits. We are particularly interested in population coding, a widespread coding scheme in the brain, in which variables are encoded through the concerted activity of large sets of neurons. Our research on spatial representations explores how the brain represents the position of objects and how these representations are used to control spatial behaviors such as reaching or navigation. We have developed a theoretical framework, based on the theory of basis functions, which accounts for the response of single cells in the parietal cortex and which explains the behavior of human patients suffering from hemineglect — a severe impairment of spatial perception. We also use experiments on normals and patients to investigate spatial memory and multisensory integration.