# Temporal Overdrive Recurrent Neural Network

## TL;DR

This paper introduces a new recurrent neural network architecture called Temporal Overdrive RNN, designed to better model systems with multiple timescales by using specialized groups of neurons for each timescale, improving system identification.

## Contribution

The novel architecture explicitly models multiple timescales within a recurrent network, enhancing its ability to predict complex time series compared to existing models.

## Key findings

- Promising results on synthetic data time series prediction
- Outperforms several state-of-the-art recurrent architectures in initial tests
- Demonstrates improved modeling of systems with multiple characteristic timescales

## Abstract

In this work we present a novel recurrent neural network architecture designed to model systems characterized by multiple characteristic timescales in their dynamics. The proposed network is composed by several recurrent groups of neurons that are trained to separately adapt to each timescale, in order to improve the system identification process. We test our framework on time series prediction tasks and we show some promising, preliminary results achieved on synthetic data. To evaluate the capabilities of our network, we compare the performance with several state-of-the-art recurrent architectures.