Evaluating the extent to which homeostatic plasticity learns to compute prediction errors in unstructured neuronal networks

TL;DR

This paper investigates whether homeostatic inhibitory synaptic plasticity enables unstructured neuronal networks to compute prediction errors, finding it effective for simple stimuli but insufficient for complex, time-varying inputs.

Contribution

The study combines simulations and mathematical analysis to demonstrate the capabilities and limitations of homeostatic plasticity in predictive coding within unstructured networks.

Findings

Homeostatic plasticity can compute prediction errors for simple stimuli.

It is insufficient for complex, time-varying stimuli.

Mean-field theory explains the conditions for effective prediction error computation.

Abstract

The brain is believed to operate in part by making predictions about sensory stimuli and encoding deviations from these predictions in the activity of "prediction error neurons." This principle defines the widely influential theory of predictive coding. The precise circuitry and plasticity mechanisms through which animals learn to compute and update their predictions are unknown. Homeostatic inhibitory synaptic plasticity is a promising mechanism for training neuronal networks to perform predictive coding. Homeostatic plasticity causes neurons to maintain a steady, baseline firing rate in response to inputs that closely match the inputs on which a network was trained, but firing rates can deviate away from this baseline in response to stimuli that are mismatched from training. We combine computer simulations and mathematical analysis systematically to test the extent to which randomly connected, unstructured networks compute prediction errors after training with homeostatic inhibitory synaptic plasticity. We find that homeostatic plasticity alone is sufficient for computing prediction errors for trivial time-constant stimuli, but not for more realistic time-varying stimuli. We use a mean-field theory of plastic networks to explain our findings and characterize the assumptions under which they apply.