Research

Our passion is to uncover principles of neural computation across species and systems. More specifically, my lab’s research is driven by two major interests:

(1) We are convinced that neurons are not just simple units whose function is to integrate inputs. Admittedly, the striking success of artificial neural networks in solving complex tasks is grounded on learning processes that adjust connectivity weights in networks of simplified neurons that sum inputs. Yet these approaches usually neglect that real neurons provide a tremendous computational repertoire of their own. The analogue, nonlinear computations performed by individual cells are highly efficient, in particular as they can be implemented at the compartmental (like the dendritic) and even the molecular level. Moreover, cellular properties are readily adapted to different states via a diverse set of neuromodulators. To date, however, it is still unclear to which extent and how cell-based computations shape network function. Our research, therefore, seeks to shed light on how cellular processes bear consequences for network behavior and contribute to neural computation.

(2) No matter which brain region or part of the nervous system we investigate, to decipher neural computation investigations primarily focus on the specific system’s computational task. From an evolutionary as well as engineering perspective, however, investments to achieve a particular function are only well placed if the latter can be maintained stably, without being easily compromised by external or intrinsic influences. This is why we consider it important to also study neural function under evolutionary constraints. These may include limited energetic resources, variable temperatures, size limitations, or a degree of flexibility that allows a system to switch between different modes. Our goal is to understand whether and how evolutionary constraints have impacted the principles of neural design.

To address these topics, we employ computational and mathematical methods. Our approach is interdisciplinary and we actively seek to engage in collaborative projects with experimental groups: be it to put theoretical predictions to test or to be inspired by experimental observations that guide new modelling approaches. We thus exploit the “luxury of the theoretician” and apply the flexible modeling and data analysis methods to a wide range of species and systems. The lab’s research work thus encompasses projects ranging from the auditory system in the locust to navigation in the mammalian hippocampal formation. The largest benefit of this wide scope is that it allows for an overarching view beyond a specific system that sharpens the eye in the search of generic principles that extend beyond a particular system.