Memory Capacity in Reservoir Computing Models

This repository contains code for papers regarding the topic of memory capacity in reservoir computing (RC) models.

List of papers with code:

• Memory capacity of nonlinear recurrent networks: Is it informative? (2025)
• Memory of recurrent networks: Do we compute it right? (2023)

---

Memory capacity of nonlinear recurrent networks: Is it informative? (2025)

Giovanni Ballarin¹, Lyudmila Grigoryeva^1,2, Juan-Pablo Ortega³

1. Faculty of Mathematics and Statistics, University of St. Gallen
2. Honorary Associate Professor, Department of Statistics, University of Warwick
3. Division of Mathematical Sciences, Nanyang Technological University

Links to paper: [ArXiv]

We provide the MATLAB code to reproduce the simulations in the paper.

Folder: mem_nonlin_recurrent_networks_2025

Memory of recurrent networks: Do we compute it right? (2023)

Giovanni Ballarin¹, Lyudmila Grigoryeva^2,3, Juan-Pablo Ortega⁴

1. Department of Economics, University of Mannheim
2. Faculty of Mathematics and Statistics, University of St. Gallen
3. Honorary Associate Professor, Department of Statistics, University of Warwick
4. Division of Mathematical Sciences, Nanyang Technological University

Links to paper: [ArXiv] [ResearchGate]

We provide the scripts necessary to reproduce all figures in the paper. All code was written in MATLAB.

Folder: mem_recurrent_networks_right_2023