The brain can be considered as a set of specialized areas for specific computations whose outputs need to be transferred to other specialized areas. Based on this principle, cognition and adaptive behavior can emerge. Unraveling how those local computations are implemented and information is transferred within the widely fixed anatomy of brain networks therefore is key for understanding cognitive processes from active perception to learning. 

​

The aim of our team’s project (ERC funded) is to test general frameworks for local computational processing (e.g. predictive processing framework) and flexible communication between brain areas based on nested oscillations (Bonnefond et al. 2017, in prep.). 

​

We use a computational, multimodal (including laminar fMRI and MEG), multi-scale and cross-species (human and monkey) approach in experiments manipulating visual processing, attention, predictability and learning to test core predictions of the frameworks. 

​

Moreover, for understanding the computational role of oscillatory brain activity, we investigate how the link between observed brain oscillations at different frequency bands relates computational mechanisms in the brain to (convolutional) Deep and Spiking Neural Networks (SNN/DNNs).