GraphiT: Encoding Graph Structure in Transformers

TL;DR

GraphiT introduces a transformer-based model that effectively encodes graph structure using relative positional encoding and local sub-structure enumeration, outperforming classical GNNs and enabling interpretability.

Contribution

The paper presents GraphiT, a novel transformer architecture that incorporates structural and positional graph information through relative encoding and local sub-structure encoding, surpassing traditional GNNs.

Findings

Outperforms classical GNNs on various tasks

Effective encoding of local graph motifs

Provides interpretable visualizations of graph features

Abstract

We show that viewing graphs as sets of node features and incorporating structural and positional information into a transformer architecture is able to outperform representations learned with classical graph neural networks (GNNs). Our model, GraphiT, encodes such information by (i) leveraging relative positional encoding strategies in self-attention scores based on positive definite kernels on graphs, and (ii) enumerating and encoding local sub-structures such as paths of short length. We thoroughly evaluate these two ideas on many classification and regression tasks, demonstrating the effectiveness of each of them independently, as well as their combination. In addition to performing well on standard benchmarks, our model also admits natural visualization mechanisms for interpreting graph motifs explaining the predictions, making it a potentially strong candidate for scientific applications where interpretation is important. Code available at https://github.com/inria-thoth/GraphiT.