The researcher will develop theories of distributed neural computation and work closely with our outstanding experimental and theoretical team members to test these theories. The two major projects will be: - Reverse-engineering the inference algorithm used by mammalian cortex, by connecting circuit architecture, distributed computation, probabilistic inference, and learning. - Revealing how brains play video games. We have designed dynamic, immersive virtual environments for animals, and will simultaneously record many neurons while the animals behave. Researchers will use control theory and network analysis to model how recorded neural activity represents amibiguous sensory variables and what neural transformations explain the animal's behavior in the games.

Requires a Ph.D. in a relevant quantitative discipline, and demonstrable expertise in computational neuroscience, statistics, machine learning, information theory, or control theory. Strong communication skills are also required. Candidates must have outstanding mathematical skills. Previous knowledge of neurobiology is advantageous but not essential.

Postdoctoral fellowships are for an initial appointment of one year, with further extensions depending on progress. The positions provide a competitive annual salary and benefits. The start date is flexible, and we look forward to reviewing completed applications as they arrive.