Abstract. This paper introduces a generative model equivariant to Euclidean symmetries: E(n) Equivariant Normalizing Flows (E-NFs). To construct E ...

In 2021, NeurIPS introduced a new track, Datasets and Benchmarks. The first year of that track, 2021, has its own proceedings, accessible by the link below.

NeurIPS 2020 · Advances in Neural... · Book

We show that diffusion models can achieve image sample quality superior to the current state-of-the-art generative models. We achieve this on unconditional ...

Advances in Neural Information Processing Systems 34 (NeurIPS 2021) ; Beyond Value-Function Gaps: Improved Instance-Dependent Regret Bounds for Episodic ...

We propose a new framework for estimating generative models via adversarial nets, in which we simultaneously train two models: a generative model G that ...

We propose a novel, simple network architecture based solely onan attention mechanism, dispensing with recurrence and convolutions entirely.Experiments on two ...

We demonstrate these properties in continuous-depth residual networks and continuous-time latent variable models. We also construct continuous normalizing flows ...