Dreaming neural networks: forgetting spurious memories and reinforcing pure ones

TL;DR

This paper introduces an extended Hopfield neural network model inspired by sleep and dreaming that unlearns spurious memories and reinforces pure ones, achieving maximal storage capacity and robustness against noise.

Contribution

The authors propose a novel sleep-inspired unlearning and consolidation mechanism that enhances the storage capacity of Hopfield networks to the theoretical maximum of alpha=1.

Findings

Achieves maximal storage capacity alpha=1.

Proves convergence of the Hebbian kernel to the pure pattern projection matrix.

Validates theoretical predictions with extensive numerical simulations.

Abstract

The standard Hopfield model for associative neural networks accounts for biological Hebbian learning and acts as the harmonic oscillator for pattern recognition, however its maximal storage capacity is $\alpha \sim 0.14$, far from the theoretical bound for symmetric networks, i.e. $\alpha =1$. Inspired by sleeping and dreaming mechanisms in mammal brains, we propose an extension of this model displaying the standard on-line (awake) learning mechanism (that allows the storage of external information in terms of patterns) and an off-line (sleep) unlearning$\&$consolidating mechanism (that allows spurious-pattern removal and pure-pattern reinforcement): this obtained daily prescription is able to saturate the theoretical bound $\alpha=1$, remaining also extremely robust against thermal noise. Both neural and synaptic features are analyzed both analytically and numerically. In particular, beyond obtaining a phase diagram for neural dynamics, we focus on synaptic plasticity and we give explicit prescriptions on the temporal evolution of the synaptic matrix. We analytically prove that our algorithm makes the Hebbian kernel converge with high probability to the projection matrix built over the pure stored patterns. Furthermore, we obtain a sharp and explicit estimate for the "sleep rate" in order to ensure such a convergence. Finally, we run extensive numerical simulations (mainly Monte Carlo sampling) to check the approximations underlying the analytical investigations (e.g., we developed the whole theory at the so called replica-symmetric level, as standard in the Amit-Gutfreund-Sompolinsky reference framework) and possible finite-size effects, finding overall full agreement with the theory.