Critical Echo State Networks that Anticipate Input using Morphable Transfer Functions

TL;DR

This paper introduces a novel critical echo state network model with adaptable neuron transfer functions that anticipate future inputs, leading to slow forgetting of deviations and potential applications in biological and technical systems.

Contribution

It presents a new critical echo state network framework with morphable transfer functions for input anticipation, supported by theoretical analysis and numerical experiments.

Findings

Deviations from expected input are forgotten in a power law manner.

The model demonstrates input anticipation capabilities.

Implications for biological and technical systems are discussed.

Abstract

The paper investigates a new type of truly critical echo state networks where individual transfer functions for every neuron can be modified to anticipate the expected next input. Deviations from expected input are only forgotten slowly in power law fashion. The paper outlines the theory, numerically analyzes a one neuron model network and finally discusses technical and also biological implications of this type of approach.