Spiking neural networks represent quantum states

Illustration of simulating restricted Boltzmann Machines with spiking neurons.
Illustration of a spiking neuron process.

Spiking neural networks emulate the behaivour of the mammal brain. Detailed numerical models allow us to simulate networks of spiking neurons on conventional computers. Furthermore, spiking neuromorphic hardware enables the efficient evaluation of evolutions in spiking neural networks on analog or mixed-signal electronic devices.

We use spiking neural networks to represent quantum many-body systems. We expect to overcome limitations of conventional computers in the simulation of quantum many-body systems by using the fast and parallel nature of spiking neural networks.

We further expect to gain new knowledge by relating the biologically inspired behaviour of spiking neurons to quantum mechanics.

Related publications