Logical Distillation of Graph Neural Networks

Pluska, Alexander and Welke, Pascal and Gärtner, Thomas and Malhotra, Sagar (2024) Logical Distillation of Graph Neural Networks. In: Proceedings of the 21st International Conference on Principles of Knowledge Representation and Reasoning :. IJCAI, pp. 920-930. ISBN 9781956792058

Text (2406.07126v3 (1))
2406.07126v3_1_.pdf - Accepted Version
Available under License Creative Commons Attribution-NonCommercial-NoDerivs.
Download (229kB)

Abstract

We present a logic based interpretable model for learning on graphs and an algorithm to distill this model from a Graph Neural Network (GNN). Recent results have shown connections between the expressivity of GNNs and the two-variable fragment of first-order logic with counting quantifiers (C2). We introduce a decision-tree based model which leverages an extension of C2 to distill interpretable logical classifiers from GNNs. We test our approach on multiple GNN architectures. The distilled models are interpretable, succinct, and attain similar accuracy to the underlying GNN. Furthermore, when the ground truth is expressible in C2, our approach outperforms the GNN.