The leaky integrator with recurrent inhibition as a predictor

TL;DR

This paper demonstrates that adding time-delayed recurrent inhibition to a leaky integrator enables it to predict certain frequency components of input signals through negative group delay, with analytical derivation and application to neural data.

Contribution

It introduces a novel predictive mechanism in leaky integrators by incorporating recurrent inhibition, revealing their potential as neural predictors.

Findings

Leaky integrator with recurrent inhibition exhibits negative group delay.

The predictive property is analytically derived.

Application demonstrated on local field potential data.

Abstract

It is shown that the leaky integrator, the basis for many neuronal models, possesses a negative group delay when a time-delayed recurrent inhibition is added to it. By means of this negative group delay, the leaky integrator becomes a predictor for some frequency components of the input signal. The prediction properties are derived analytically and an application to a local field potential is provided.