Logical Distillation of Graph Neural Networks

Alexander Pluska, Pascal Welke, Thomas Gärtner, Sagar Malhotra
6/11/2024
cs.LGcs.AI

No official code found. Best implementation: lammdachs/LogicalDistillationOfGNNs (2 stars, last commit 2024-05-28).

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.

View on arXivView PDF

Code Implementations(1)

lammdachs/LogicalDistillationOfGNNsCommunity100%
21Shell, PythonFeb 29, 20242 years ago

Similar Papers

Neural Logic Networks for Interpretable Classification

85% match2025

Neural Logic Networks for Interpretable Classification

85% match2025

Calibrate and Boost Logical Expressiveness of GNN Over Multi-Relational and Temporal Graphs

82% match2026

Efficiently Learning Probabilistic Logical Models by Cheaply Ranking Mined Rules

81% match2024

Learning Interpretable Differentiable Logic Networks

80% match2 repos12024

On the Limits of Hierarchically Embedded Logic in Classical Neural Networks

80% match2025

Cite this paper

@article{pluska2024logical,
  title  = {Logical Distillation of Graph Neural Networks},
  author = {Alexander Pluska and Pascal Welke and Thomas Gärtner and Sagar Malhotra},
  year   = {2024},
  eprint = {2406.07126},
  archivePrefix = {arXiv},
  url    = {http://arxiv.org/abs/2406.07126v3}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/2406.07126.svg)](https://codeofpaper.com/paper/2406.07126)

Discussion