Alexandre Pouget is currently a full professor in the Department of Basic Neuroscience at the University of Geneva where he leads the computational cognitive neuroscience laboratory. His research focuses on general theories of representation and computation in neural circuits with a strong emphasis on neural theories of probabilistic inference. This approach is built on the notion that knowledge in the brain takes the form of probability distributions, and new knowledge is acquired via probabilistic inference. This allows robust computations in the presence of uncertainty, a situation that arises in almost all real-life computations. He is currently applying this framework to a wide range of topics including olfactory processing, spatial representations, sensory-motor transformations, multisensory integration, perceptual learning, attention control, decision making, causal reasoning and simple arithmetic.
The CNBC will present the award to Pouget at 4 P.M. on Wednesday, May 11, 2016, at the Rashid auditorium in CMU’s Hillman Center for Future Generation Technologies.
ABSTRACT: Most computations performed by the brain are subject to a large amount of uncertainty because sensory inputs are noisy and ambiguous. As Laplace and Hemlholtz have pointed out over the last two centuries, the best way to compute in the presence of uncertainty is to adopt a probabilistic approach, that is, to represent knowledge in the form of probability distributions and to perform probabilistic computations. Several hypotheses have emerged recently regarding the neural implementations of these probabilistic inferences. I will review in particular one such hypothesis, based on the notion of probabilistic population codes, which shows how neurons perform probabilistic inference with simple biologically plausible linear and nonlinear neural circuits. We are applying this approach to a wide array of seemingly different behaviours such as decision making, visual search, simple arithmetic, perceptual learning, multisensory integration and olfactory processing to name a few. Interestingly, the mechanisms involved are so simple that they might also be used in insects, suggesting that probabilistic inference provides a general framework to understand neural computation in all species.