The computational and learning benefits of Daleian neural networks

Abstract

Dale's principle implies that biological neural networks are composed of neurons that are either excitatory or inhibitory. While the number of possible architectures of such Daleian networks is exponentially smaller than non-Daleian ones, the computational and functional implications of using Daleian networks by the brain are mostly unknown. Here, we use models of recurrent spiking neural networks and rate-based networks to show, surprisingly, that despite the structural limitations on Daleian networks, they can approximate the computation performed by non-Daleian networks to a very high degree of accuracy. Moreover, we find that Daleian networks are more functionally robust to synaptic noise. We then show that unlike non-Daleian networks, Daleian ones can learn efficiently by tuning single neuron features, nearly as well as learning by tuning individual synaptic weightssuggesting a simpler and more biologically plausible learning mechanism. We thus suggest that in addition to architectural simplicity, Dale's principle confers computational and learning benefits for biological networks, and offers new directions for constructing and training biologically-inspired artificial neural networks.36th Conference on Neural Information Processing Systems (NeurIPS 2022).

Homeostatic synaptic normalization optimizes learning in network models of neural population codes

“…Moreover, learning such network models would be of key interest, both as a potential way to improve learning in artificial neural networks using biological features, and as a way for biological neural networks to implement efficient learning and overcome the credit assignment problem (33)(34)(35)(36)(37)(38). Both structured architectural features of neural circuits and random connectivity patterns have been suggested to enable or shape the computation carried out by neural circuits (30,(39)(40)(41)(42)(43). In particular, these computations rely on the nature of synaptic connectivity and the coupling between synapses in terms of how they change during learning.…”

“…In addition, they may be useful for improving learning in artificial neural networks using biological features (33)(34)(35)(36)(37)(38). Both structured architectural features of neural circuits and random connectivity patterns have been suggested to shape the computation carried out by neural circuits (30,(39)(40)(41)(42)(43). These computations rely on the nature of synaptic connectivity and the coupling between synapses in terms of how they change during learning.…”