Logical Distillation of Graph Neural Networks | Papers With Code

Retrieved on: 2024-06-12 18:00:54

Tags for this article:

Model theory

Predicate logic

Artificial neural networks

Graph neural network

Semi-supervised learning

Descriptive complexity theory

Click the tags to see associated articles and topics

Logical Distillation of Graph Neural Networks | Papers With Code. View article details on hiswai:

Summary

The article introduces a logic-based interpretable model distillation method for Graph Neural Networks (GNNs), leveraging two-variable logic with counting quantifiers (C2). Tags like 'Model theory' and 'Descriptive complexity theory' contextualize it within interpretability in AI and logic.

Article found on: paperswithcode.com

View Original Article