Welcome to Guillaume Dehaene’s blog.

I’m a young PhD student at university of Geneva, where I study computational neuroscience. I’m also quite interested in Bayesian statistics, and I’m hoping to find a post-doc position in that field.

I have done some research on the subject of understanding the Expectation Propagation method for approximate inference. On the way there, I have developed some results on log-concave distributions, which I believe can be of great use to the theoretical understanding of why some Bayesian methods just work way better than others.