Shuffling Recurrent Neural Networks

Authors

  • Michael Rotman Tel Aviv University, Israel
  • Lior Wolf Tel Aviv University, Israel

Keywords:

(Deep) Neural Network Algorithms

Abstract

We propose a novel recurrent neural network model, where the hidden state hₜ is obtained by permuting the vector elements of the previous hidden state hₜ₋₁ and adding the output of a learned function β(xₜ) of the input xₜ at time t. In our model, the prediction is given by a second learned function, which is applied to the hidden state s(hₜ). The method is easy to implement, extremely efficient, and does not suffer from vanishing nor exploding gradients. In an extensive set of experiments, the method shows competitive results, in comparison to the leading literature baselines. We share our implementation at https://github.com/rotmanmi/SRNN.

Downloads

Published

2021-05-18

How to Cite

Rotman, M., & Wolf, L. (2021). Shuffling Recurrent Neural Networks. Proceedings of the AAAI Conference on Artificial Intelligence, 35(11), 9428-9435. Retrieved from https://ojs.aaai.org/index.php/AAAI/article/view/17136

Issue

Section

AAAI Technical Track on Machine Learning IV