A Differential Geometry Perspective on Orthogonal Recurrent Models

Abstract

Recently, orthogonal recurrent neural networks (RNNs) have emerged as state-of-the-art models for learning long-term dependencies. This class of models mitigates the exploding and vanishing gradients problem by design. In this work, we employ tools and insights from differential geometry to offer a novel perspective on orthogonal RNNs. We show that orthogonal RNNs may be viewed as optimizing in the space of divergence-free vector fields. Specifically, based on a well-known result in differential geometry that relates vector fields and linear operators, we prove that every divergence-free vector field is related to a skew-symmetric matrix. Motivated by this observation, we study a new recurrent model, which spans the entire space of vector fields. Our method parameterizes vector fields via the directional derivatives of scalar functions. This requires the construction of latent inner product, gradient, and divergence operators. In comparison to state-of-the-art orthogonal RNNs, our approach achieves comparable or better results on a variety of benchmark tasks.

References in corpus (7)

• On the difficulty of training Recurrent Neural Networks
• Modeling Temporal Dependencies in High-Dimensional Sequences: Application to Polyphonic Music Generation and Transcription
• Full-Capacity Unitary Recurrent Neural Networks
• AntisymmetricRNN: A Dynamical System View on Recurrent Neural Networks
• Physics-informed Autoencoders for Lyapunov-stable Fluid Flow Prediction
• Continuous-in-Depth Neural Networks
• Towards Robust ResNet: A Small Step but A Giant Leap