We Cannot Guarantee Safety: The Undecidability of Graph Neural Network Verification

by   Marco Sälzer, et al.
Universität Kassel

Graph Neural Networks (GNN) are commonly used for two tasks: (whole) graph classification and node classification. We formally introduce generically formulated decision problems for both tasks, corresponding to the following pattern: given a GNN, some specification of valid inputs, and some specification of valid outputs, decide whether there is a valid input satisfying the output specification. We then prove that graph classifier verification is undecidable in general, implying that there cannot be an algorithm surely guaranteeing the absence of misclassification of any kind. Additionally, we show that verification in the node classification case becomes decidable as soon as we restrict the degree of the considered graphs. Furthermore, we discuss possible changes to these results depending on the considered GNN model and specifications.


page 1

page 2

page 3

page 4


Non-IID Graph Neural Networks

Graph classification is an important task on graph-structured data with ...

edGNN: a Simple and Powerful GNN for Directed Labeled Graphs

The ability of a graph neural network (GNN) to leverage both the graph t...

Watermarking Graph Neural Networks by Random Graphs

Many learning tasks require us to deal with graph data which contains ri...

Neural Network Branch-and-Bound for Neural Network Verification

Many available formal verification methods have been shown to be instanc...

Algorithm Selection for Software Verification using Graph Attention Networks

The field of software verification has produced a wide array of algorith...

Neural Network Branching for Neural Network Verification

Formal verification of neural networks is essential for their deployment...

What graph neural networks cannot learn: depth vs width

This paper studies the capacity limits of graph neural networks (GNN). R...

Please sign up or login with your details

Forgot password? Click here to reset