Mental Illness as impaired Bayesian Inference

A growing idea in computational neuroscience is that the brain can be
viewed as a sort of "guessing machine", constantly trying to guess
what is present in the external world, what is the best action to take
and automatically trying to predict the next moment.
The way the brain would do that is by maintaining and updating
internal probabilistic models of the world that serve to interpret the
environment and guide our actions, and using calculations akin to the
well known statistical methods of "Bayesian inference".
This idea is increasingly recognised to also be of interest to Psychiatry.
Mental illness could correspond to the brain trying to interpret the
world through distorted internal models, or incorrectly combining such
internal models with sensory information.

I will describe work pursued in my lab that aims at uncovering such
internal models, using behavioural experiments and computational
methods. In health, we are particularly interested in clarifying how
prior beliefs affect perception and decision-making, how long they
take to build up or be unlearned, how complex they can be, and how
they can inform us on the type of computations and learning that the
brain performs. In mental illness, we are interested in understanding
whether/how the machinery of probabilistic inference could be
impaired, and/or relies on the use of distorted prior beliefs.
I will describe recent results relevant to the study of Schizophrenia,
Autism and Depression.