Computational Neurophysics

The cortical neuronal network is among the most complex structures found in nature. The functional role of its dynamics exhibited on many spatio-temporal scales is presently not understood. Furthermore, in contrast to other systems, the structure of the cortex is in fact not static but undergoes a continuous activity dependent reorganization. The Diesmann Research Unit studies the mechanisms and functional consequences of spike synchronization and plasticity in biologically realistic models of the cortical network. However, this bottom-up approach alone may not lead to an understanding of brain function. For this reason we also incorporate top-down approaches in our research. At the interface of top-down and bottom up approaches, our strategy is to implement established formal theories of system function like temporal-difference learning in biologically constrained network models. These investigations depend on large-scale simulations requiring non-standard algorithms and high-performance parallel computing. Therefore, the unit is also concerned with the creation of appropriate simulation technology.