My research explores how the mind copes with the uncertainty that pervades perception and cognition. My working hypothesis is that the brain performs large-scale probabilistic inference in parallel over a rich set of internal variables  that represent the hidden structure of the environment. Such an approach would be particularly well suited to accumulating information across successive observations and generalising knowledge across partially related tasks. I investigate this hypothesis by analysing behavioural and neural data, combining ideal-observer analyses with hierarchical Bayesian modelling.

The very richness that makes this account powerful also creates a computational challenge: probabilistic inference over a rich internal model quickly becomes intractable. This suggests that the mind may need to selectively recruit background knowledge for inference in a particular situation, balancing inferential accuracy against computational tractability. More recently, I have started to engage with the question of how this selection might be achieved.

Qualifications:

BSc: Physics, Budapest University of Technology and Economics (Budapest, 2010)

MSc: Physics, Eotvos Lorand University (Budapest, 2013)