Learning Universal Computations with Spikes

Providing the neurobiological basis of information processing in higher animals, spiking neural networks must be able to learn a variety of complicated computations, including the generation of appropriate, possibly delayed reactions to inputs and the self-sustained generation of complex activity patterns, e.g. for locomotion. Many such computations require previous building of intrinsic world models. Here we show how spiking neural networks may solve these different tasks. Firstly, we derive constraints under which classes of spiking neural networks lend themselves to substrates of powerful general purpose computing. The networks contain dendritic or synaptic nonlinearities and have a constrained connectivity. We then combine such networks with learning rules for outputs or recurrent connections. We show that this allows to learn even difficult benchmark tasks such as the self-sustained generation of desired low-dimensional chaotic dynamics or memory-dependent computations. Furthermore, we show how spiking networks can build models of external world systems and use the acquired knowledge to control them.Author Summary: Animals and humans can learn versatile computations such as the generation of complicated activity patterns to steer movements or the generation of appropriate outputs in response to inputs. Such learning must be accomplished by networks of nerve cells in the brain, which communicate with short electrical impulses, so-called spikes. Here we show how such networks may perform the learning. We track their ability back to experimentally found nonlinearities in the couplings between nerve cells and to a network connectivity that complies with constraints. We show that the spiking networks are able to learn difficult tasks such as the generation of desired chaotic activity and the prediction of the impact of actions on the environment. The latter allows to compute optimal actions by mental exploration.