Summary of Goat: Explaining Graph Neural Networks Via Graph Output Attribution, by Shengyao Lu et al.
GOAt: Explaining Graph Neural Networks via Graph Output Attribution
by Shengyao Lu, Keith G. Mills, Jiao He, Bang Liu, Di Niu
First submitted to arxiv on: 26 Jan 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: None
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary The paper introduces Graph Output Attribution (GOAt), a novel method for explaining the decision-making process of Graph Neural Networks (GNNs). The approach attributes graph outputs to input graph features, providing faithful, discriminative, and stable explanations. By expanding the GNN as a sum of scalar products, the authors propose an efficient analytical method to compute the contributions of node and edge features to each output. Experimental results on synthetic and real-world data demonstrate that GOAt outperforms state-of-the-art GNN explainers in terms of fidelity, discriminability, and stability. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary GNNs are powerful tools for understanding complex networks. But how do they make decisions? That’s what researchers want to know. They developed a new way called Graph Output Attribution (GOAt) that helps understand how GNNs work. It’s like explaining a mysterious box by showing how each piece inside contributes to the output. The method is special because it works well on many types of networks and data. |
Keywords
* Artificial intelligence * Gnn
