Residual Reservoir Memory Networks

Matteo Pinna, Andrea Ceni, Claudio Gallicchio
8/13/2025
cs.LGcs.AI

No official code found. Best implementation: NennoMP/residualrmn (1 star, MIT, last commit 2026-02-07).

Abstract

We introduce a novel class of untrained Recurrent Neural Networks (RNNs) within the Reservoir Computing (RC) paradigm, called Residual Reservoir Memory Networks (ResRMNs). ResRMN combines a linear memory reservoir with a non-linear reservoir, where the latter is based on residual orthogonal connections along the temporal dimension for enhanced long-term propagation of the input. The resulting reservoir state dynamics are studied through the lens of linear stability analysis, and we investigate diverse configurations for the temporal residual connections. The proposed approach is empirically assessed on time-series and pixel-level 1-D classification tasks. Our experimental results highlight the advantages of the proposed approach over other conventional RC models.

View on arXivView PDF

Code Implementations(1)

NennoMP/residualrmnCommunity65%
MIT

[IJCNN-2025] Residual Reservoir Memory Networks

11Shell, PythonApr 30, 20258 months agoMIT
deep-learningecho-state-networksrecurrent-neural-networksreservoir-computing

Similar Papers

Deep Residual Echo State Networks: exploring residual orthogonal connections in untrained Recurrent Neural Networks

89% match2025

Reservoir Computing: A New Paradigm for Neural Networks

84% match2025

Universal Approximation of Linear Time-Invariant (LTI) Systems through RNNs: Power of Randomness in Reservoir Computing

84% match2023

Reservoir Computing for Fast, Simplified Reinforcement Learning on Memory Tasks

83% match2024

Sequential Reservoir Computing for Efficient High-Dimensional Spatiotemporal Forecasting

83% match2026

Universality of Real Minimal Complexity Reservoir

83% match1 repo2024

Cite this paper

@article{pinna2025residual,
  title  = {Residual Reservoir Memory Networks},
  author = {Matteo Pinna and Andrea Ceni and Claudio Gallicchio},
  year   = {2025},
  eprint = {2508.09925},
  archivePrefix = {arXiv},
  url    = {http://arxiv.org/abs/2508.09925v1}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/2508.09925.svg)](https://codeofpaper.com/paper/2508.09925)

Discussion