Neural Ordinary Differential Equations on Manifolds

Normalizing flows are a powerful technique for obtaining reparameterizable samples from complex multimodal distributions. Unfortunately current approaches fall short when the underlying space has a non trivial topology, and are only available for the most basic geometries. Recently normalizing flows in Euclidean space based on Neural ODEs show great promise, yet suffer the same limitations. Using ideas from differential geometry and geometric control theory, we describe how neural ODEs can be extended to smooth manifolds. We show how vector fields provide a general framework for parameterizing a flexible class of invertible mapping on these spaces and we illustrate how gradient based learning can be performed. As a result we define a general methodology for building normalizing flows on manifolds.

[1]  M. L. Chambers The Mathematical Theory of Optimal Processes , 1965 .

[2]  Adam M. Oberman,et al.  How to Train Your Neural ODE: the World of Jacobian and Kinetic Regularization , 2020, ICML.

[3]  Bohua Zhan,et al.  Smooth Manifolds , 2021, Arch. Formal Proofs.

[4]  U. Boscain,et al.  A Comprehensive Introduction to Sub-Riemannian Geometry , 2019 .

[5]  Eric Nalisnick,et al.  Normalizing Flows for Probabilistic Modeling and Inference , 2019, J. Mach. Learn. Res..

[6]  Shoichiro Yamaguchi,et al.  A Wrapped Normal Distribution on Hyperbolic Space for Gradient-Based Learning , 2019, ICML.

[7]  David Duvenaud,et al.  Neural Ordinary Differential Equations , 2018, NeurIPS.

[8]  Jimmy Ba,et al.  Adam: A Method for Stochastic Optimization , 2014, ICLR.

[9]  Gurtej Kanwar,et al.  Normalizing Flows on Tori and Spheres , 2020, ICML.

[10]  Renjie Liao,et al.  Latent Variable Modelling with Hyperbolic Normalizing Flows , 2020, ICML.

[11]  Salma A. Khalil,et al.  Types of Derivatives: Concepts and Applications (II) , 2017 .

[12]  Ana Cannas da Silva,et al.  Lectures on Symplectic Geometry , 2008 .

[13]  Shakir Mohamed,et al.  Normalizing Flows on Riemannian Manifolds , 2016, ArXiv.

[14]  Will Grathwohl Scalable Reversible Generative Models with Free-form Continuous Dynamics , 2018 .

[15]  Nicola De Cao,et al.  Hyperspherical Variational Auto-Encoders , 2018, UAI 2018.

[16]  L. Verstraelen,et al.  Laplace Transformations of Submanifolds , 2013, 1307.1515.

[17]  Ernst Hairer,et al.  Solving Differential Equations on Manifolds , 2011 .

[18]  Vlado Menkovski,et al.  Diffusion Variational Autoencoders , 2019, IJCAI.

[19]  G. Walschap Metric Structures in Differential Geometry , 2004 .

[20]  Shakir Mohamed,et al.  Variational Inference with Normalizing Flows , 2015, ICML.

[21]  Nicola De Cao,et al.  Explorations in Homeomorphic Variational Auto-Encoding , 2018, ArXiv.

[22]  Eldad Haber,et al.  Stable architectures for deep neural networks , 2017, ArXiv.

[23]  A. Agrachev,et al.  Control Theory from the Geometric Viewpoint , 2004 .