We believe that tractable models and quantitative methods drive progress
We believe that the development of new models and analysis techniques that draw on mathematics, physics and computer science knowledge drive our understanding of biology. We believe that a good model or a good analysis comes with a derivation from first principles and that it has a range of validity within which it makes accurate predictions and postdictions. A biological model can operate at different levels of abstraction, from micro scale of protein interaction to the macroscale of network dynamics. We concentrate on the network level and build models that zoom-in on the question how neurons encode incoming information and how they organize their global network dynamics to achieve this task. While we specialize in neuronal signal processing, we strive to make our models and methods applicable for other disciplines and hope to impact applications such as machine learning and brain-computer interfaces.
The role of computational models
The study of the brain was historically a biology-dominated discipline where experiments took center stage. Even though we are theorists at heart we place an emphasis on verifying our models with experimetal data. At the Max Planck Institute for Brain Research we are surrounded by excellent experimental groups which make this theory-experiment feedback-loop enjoyable and fun. When you visit us, you will walk past long corridors filled with experimental set-ups. You know you found the "Theory of neural dynamics group" when you spot the whiteboards full of equations.