Graph Normalizing Flows

By Jenny Liu, Aviral Kumar, Jimmy Ba, Jamie Ryan Kiros, Kevin Swersky. 2019

187 citations recorded by the source corpus.

We introduce graph normalizing flows: a new, reversible graph neural network model for prediction and generation. On supervised tasks, graph normalizing flows perform similarly to message passing neural networks, but at a significantly reduced memory footprint, allowing them to scale to larger graphs. In the unsupervised case, we combine graph normalizing flows with a novel graph auto-encoder to create a generative model of graph structures. Our model is permutation-invariant, generating entire graphs with a single feed-forward pass, and achieves competitive results with the state-of-the art auto-regressive models, while being better suited to parallel computing architectures.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition, Generative Models, Graph Learning

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Computer Vision and Pattern Recognition

In Generative Models

In Graph Learning

Citations in the corpus

E(n) Equivariant Graph Neural Networks (2021, methodology): “The decoder g(·) proposed by (Liu et al., 2019) takes as input the embedding space z and outputs the reconstructed adjacency matrix Â = g(z), this decoder function is defined as follows:”
On the Binding Problem in Artificial Neural Networks (2020, background): “Recently, MPNNs were extended to allow for continuous updates (Deng et al., 2019; Liu et al., 2019).”
Equivariant Flows: exact likelihood generative learning for symmetric densities (2020, methodology): “Symmetries in flow models have been discussed in the context of permutations in graphs (Liu et al., 2019a).”
Permutation Invariant Graph Generation via Score-Based Generative Modeling (2020, methodology): “In contrast, GNF is a permutation invariant model.”
MoFlow: An Invertible Flow Model for Generating Molecular Graphs (2020, background): “Examples include NICE[7], RealNVP[8], Glow[14] and GNF [18] which show promising results in generating images or even graphs [18].”
Continuous Graph Flow (2019, methodology): “We evaluate our model on graph generation on two benchmark datasets EGO - SMALL and COMMUNITY - SMALL against four state-of-the-art base-lines: GraphVAE (Simonovsky & Komodakis, 2018), GraphRNN (You et al., 2018), DeepGMG (Li et al., 2018), and Graph normalizing ﬂows (GNF) (Liu et al., 2019).”
Score-based Generative Modeling of Graphs via the System of Stochastic Differential Equations (2022, methodology): “Among them, autoregressive models (You et al., 2018b; Liao et al., 2019) construct a graph via sequential decisions, while one-shot generative models (De Cao & Kipf, 2018; Liu et al., 2019) generate components of a graph at once.”
Training Graph Neural Networks with 1000 Layers (2021, unknown): “Inspired by reversible networks (Gomez et al., 2017; Liu et al., 2019; Kitaev et al., 2019) and grouped convolutions (Krizhevsky et al.”
GraphAF: a Flow-based Autoregressive Model for Molecular Graph Generation (2020, methodology): “GNF defines a normalizing flow from a base distribution to the hidden node representations of a pretrained Graph Autoencoders.”