Title: Uncertainty in the predictive brain

Abstract: An agent striving to deal with new information optimally needs to update not only the content of its beliefs but also its uncertainty about them. We may therefore expect evolved living agents, such as humans, to update uncertainty about their beliefs with each piece of new information, and we may expect to obtain superior models of learning and decision-making by incorporating uncertainty updates in our modelling. Furthermore, we may expect to see brain responses to new information that reflect uncertainty updates. In this talk, I will first give a brief overview of the theory behind uncertainty updating and introduce recent advances in modelling uncertainty. I will then present a series of recent human behavioural and neuroimaging studies which reveal signatures of uncertainty in both healthy and psychiatric patient populations. These results underscore the importance of the computational modelling of uncertainty in learning and decision-making.

ALL ARE WELCOME

Links to his research:
https://chrismathys.com

https://scholar.google.com/citations?user=9kd9A48AAAAJ&hl=en

Link to his computational modelling toolbox:

https://www.tnu.ethz.ch/en/software/tapas/documentations/hgf-toolbox-v40.html