AntisymmetricRNN: A Dynamical System View on Recurrent Neural Networks
arXiv:1902.09689
Abstract
Recurrent neural networks have gained widespread use in modeling sequential data. Learning long-term dependencies using these models remains difficult though, due to exploding or vanishing gradients. In this paper, we draw connections between recurrent networks and ordinary differential equations. A special form of recurrent networks called the AntisymmetricRNN is proposed under this theoretical framework, which is able to capture long-term dependencies thanks to the stability property of its underlying differential equation. Existing approaches to improving RNN trainability often incur significant computation overhead. In comparison, AntisymmetricRNN achieves the same goal by design. We showcase the advantage of this new architecture through extensive simulations and experiments. AntisymmetricRNN exhibits much more predictable dynamics. It outperforms regular LSTM models on tasks requiring long-term memory and matches the performance on tasks where short-term dependencies dominate despite being much simpler.
Published as a conference paper at ICLR 2019
Cited by in corpus (9)
- Combining Recurrent, Convolutional, and Continuous-time Models with Linear State-Space Layers
- Continuous-in-Depth Neural Networks
- Dynamical Isometry and a Mean Field Theory of LSTMs and GRUs
- Framing RNN as a kernel method: A neural ODE approach
- Interpolation between Residual and Non-Residual Networks
- A Differential Geometry Perspective on Orthogonal Recurrent Models
- How Chaotic Are Recurrent Neural Networks?
- Online Spatiotemporal Action Detection and Prediction via Causal Representations
- Review: Ordinary Differential Equations For Deep Learning