In previous work, we introduced the “Josephson junction neuron,” a superconducting circuit that can reproduce some important properties of biological neurons like action potentials and refractory periods. Circuits of Josephson junction neurons have very fast dynamics and operate fully in parallel, and large simulations would be possible with them that are not feasible using traditional computers. Here, we show that this model reproduces 15 common dynamical behaviors of individual cortical spiking neurons for various choices of the intrinsic circuit parameters, while 4 more behaviors can be produced by also including a sinusoidal bias current, giving a total of 19. Behaviors such as single spike response, tonic (periodic) spiking only when stimulated, bursting, tonic bursting, frequency adaptation, and subthreshold oscillations are some of the common behaviors seen. Even bistability of periodic firing and quiescent modes are possible. This suggests that Josephson junction neurons could be fabricated and used to simulate biologically realistic neural circuits, networks, and perhaps even larger regions of the brain.
Read full abstract