FFJORD: Free-form continuous dynamics for scalable reversible generative models
Captured source
source ↗FFJORD: Free-form continuous dynamics for scalable reversible generative models | OpenAI
October 2, 2018
Publication
FFJORD: Free-form continuous dynamics for scalable reversible generative models
Read paper
Loading…
Share
Abstract
A promising class of generative models maps points from a simple distribution to a complex distribution through an invertible neural network. Likelihood-based training of these models requires restricting their architectures to allow cheap computation of Jacobian determinants. Alternatively, the Jacobian trace can be used if the transformation is specified by an ordinary differential equation. In this paper, we use Hutchinson's trace estimator to give a scalable unbiased estimate of the log-density. The result is a continuous-time invertible generative model with unbiased density estimation and one-pass sampling, while allowing unrestricted neural network architectures. We demonstrate our approach on high-dimensional density estimation, image generation, and variational inference, achieving the state-of-the-art among exact likelihood methods with efficient sampling.
- Generative Models
Authors
Will Grathwohl, Ricky T. Q. Chen, Jesse Bettencourt, Ilya Sutskever, David Duvenaud
Related articles
View all
Hierarchical text-conditional image generation with CLIP latentsPublicationApr 13, 2022
DALL·E: Creating images from textMilestoneJan 5, 2021
Image GPTPublicationJun 17, 2020