Timezone: »

Residual Flows for Invertible Generative Modeling
Tian Qi Chen · Jens Behrmann · David Duvenaud · Joern-Henrik Jacobsen

Tue Dec 10 04:40 PM -- 04:45 PM (PST) @ West Exhibition Hall C + B3

Flow-based generative models parameterize probability distributions through an invertible transformation and can be trained by maximum likelihood. Invertible residual networks provide a flexible family of transformations where only Lipschitz conditions rather than strict architectural constraints are needed for enforcing invertibility. However, prior work trained invertible residual networks for density estimation by relying on biased log-density estimates whose bias increased with the network's expressiveness. We give a tractable unbiased estimate of the log density, and reduce the memory required during training by a factor of ten. Furthermore, we improve invertible residual blocks by proposing the use of activation functions that avoid gradient saturation and generalizing the Lipschitz condition to induced mixed norms. The resulting approach, called Residual Flows, achieves state-of-the-art performance on density estimation amongst flow-based models, and outperforms networks that use coupling blocks at joint generative and discriminative modeling.

Author Information

Tian Qi Chen (U of Toronto)
Jens Behrmann (University of Bremen)
David Duvenaud (University of Toronto)

David Duvenaud is an assistant professor in computer science at the University of Toronto. His research focuses on continuous-time models, latent-variable models, and deep learning. His postdoc was done at Harvard University, and his Ph.D. at the University of Cambridge. David also co-founded Invenia, an energy forecasting and trading company.

Joern-Henrik Jacobsen (Vector Institute)

Related Events (a corresponding poster, oral, or spotlight)

More from the Same Authors