Our investigations take place in the field of Computational Neuroscience.  To understand neural computation, we believe strongly it is first necessary to determine the dynamical properties of the neural elements involved. It is clear from recent work that single neurons are already very complicated non-linear devices, as exemplified by the cerebellar Purkinje cell [references], and their intrinsic properties may contribute significantly to the computation taking place in neural networks. Our present work therefore consists of a bottom-up approach to examine neural dynamics in cerebellar and basal ganglia circuits. Most importantly, we focus on the question of how thousands of synaptic inputs that a single neurons receives per second in a behaving animal may control the output spiking of this neuron. This control of spiking is to a large degree mediated by the activation of voltage-gated conductances and is generally not a simple linear transform of the input spike rates.

One interesting side issue that is important to our work is the question of how inhibitory inputs are involved in neural computation. Both cerebellar and basal ganglia circuits are somewhat unusual in the brain in that inhibitory pathways convey the main stream of information between structures. Since inhibition largely reduces spike output, these circuits likely convey information through brief pauses in spiking. Clear evidence [references] for this mechanism has been found in the basal ganglia circuit controlling eye movement. This is different from cerebral cortex, where inhibition is solely exerted by local interneurons, and usually has been assigned an auxiliary functional role.