I am interested in the insect olfactory system as a stereotype model for neural computation. My main focus is on odor representation in Drosophila and the Honeybee at different levels of olfactory processing such as the antennal lobe and the mushroom body. The key questions comprise how the input representation interplays with neuronal variability. What is optimal in terms robust learning and memory traces? My approach includes theoretical models and simulation of large scale, realistic spiking neuronal networks.
My current project is aimed towards understanding how the connectivity structure and typical characteristics of neuronal networks, interact and constrain neural computation, i.e. the representation of a sensory input. The approach to decouple the interactions, opens a perspective on a new class of computational paradims in structure-dynamics interaction. In particular, in a model of the insect olfactory system, I am interested in using this approach to understand odor invoked spatio-temporal patterns and established a link to a odor memory trace.