Recipe for a General, Powerful, Scalable Graph Transformer
By Ladislav Rampášek, Mikhail Galkin, Vijay Prakash Dwivedi, Anh Tuan Luu, Guy Wolf, Dominique Beaini. Neural Information Processing Systems, 2022
1,075 citations recorded by the source corpus.
We propose a recipe on how to build a general, powerful, scalable (GPS) graph Transformer with linear complexity and state-of-the-art results on a diverse set of benchmarks. Graph Transformers (GTs) have gained popularity in the field of graph representation learning with a variety of recent publications but they lack a common foundation about what constitutes a good positional or structural encoding, and what differentiates them. In this paper, we summarize the different types of encodings with a clearer definition and categorize them as being $\textit{local}$, $\textit{global}$ or $\textit{relative}$. The prior GTs are constrained to small graphs with a few hundred nodes, here we propose the first architecture with a complexity linear in the number of nodes and edges $O(N+E)$ by decoupling the local real-edge aggregation from the fully-connected Transformer. We argue that this decoupling does not negatively affect the expressivity, with our architecture being a universal function approximator on graphs. Our GPS recipe consists of choosing 3 main ingredients: (i) positional/structural encoding, (ii) local message-passing mechanism, and (iii) global attention mechanism. We provide a modular framework $\textit{GraphGPS}$ that supports multiple types of encodings and that provides efficiency and scalability both in small and large graphs. We test our architecture on 16 benchmarks and show highly competitive results in all of them, show-casing the empirical benefits gained by the modularity and the combination of different strategies.
Subfields: Artificial Intelligence, Management Science and Operations Research
View Paper: https://www.semanticscholar.org/paper/277dd73bfeb5c46513ce305136b0e71fcd2a311c
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 Management Science and Operations Research
Citations in the corpus
- Large Language Models on Graphs: A Comprehensive Survey (2023, methodology): “…typically adopt the absolute or relative positional encoding considering the position of the word token in the sequence, while graph Transformers adopt shortest path distance [141], random walk distance, the eigenvalues of the graph Laplacian [142] to consider the distance of nodes in the graph.”
- Long Range Graph Benchmark (2022, neutral): “Since MP-GNNs rely on information aggregated from a local neighborhood to update a node’s features, it is subject to miss global structural information, such as global positional encoding (PE) [51].”
- A Tutorial on Fluid Antenna System for 6G Networks: Encompassing Communication Theory, Optimization Methods and Hardware Designs (2024, background): “9, the modules with inductive bias for capturing local correlations and smoothness include Convolutional Neural Networks (CNN) (e.g., ResNet [132]), Local Attention (e.g., Swin [133]), Graph Neural Networks (GNN) (e.g., GPS [134]), and etc.”
- LightRAG: Simple and Fast Retrieval-Augmented Generation (2024, neutral): “Graphs are particularly effective at representing the interdependencies among different entities (Rampášek et al., 2022), which enables a more nuanced understanding of re-10746 lationships.”