Reconsidering Faithfulness in Regular, Self-Explainable and Domain Invariant GNNs

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Azzolin, Steve, Longa, Antonio, Teso, Stefano, Passerini, Andrea
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866917981612670976
author Azzolin, Steve
Longa, Antonio
Teso, Stefano
Passerini, Andrea
author_facet Azzolin, Steve
Longa, Antonio
Teso, Stefano
Passerini, Andrea
contents As Graph Neural Networks (GNNs) become more pervasive, it becomes paramount to build reliable tools for explaining their predictions. A core desideratum is that explanations are \textit{faithful}, \ie that they portray an accurate picture of the GNN's reasoning process. However, a number of different faithfulness metrics exist, begging the question of what is faithfulness exactly and how to achieve it. We make three key contributions. We begin by showing that \textit{existing metrics are not interchangeable} -- \ie explanations attaining high faithfulness according to one metric may be unfaithful according to others -- and can systematically ignore important properties of explanations. We proceed to show that, surprisingly, \textit{optimizing for faithfulness is not always a sensible design goal}. Specifically, we prove that for injective regular GNN architectures, perfectly faithful explanations are completely uninformative. This does not apply to modular GNNs, such as self-explainable and domain-invariant architectures, prompting us to study the relationship between architectural choices and faithfulness. Finally, we show that \textit{faithfulness is tightly linked to out-of-distribution generalization}, in that simply ensuring that a GNN can correctly recognize the domain-invariant subgraph, as prescribed by the literature, does not guarantee that it is invariant unless this subgraph is also faithful.The code is publicly available on GitHub
format Preprint
id arxiv_https___arxiv_org_abs_2406_15156
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Reconsidering Faithfulness in Regular, Self-Explainable and Domain Invariant GNNs
Azzolin, Steve
Longa, Antonio
Teso, Stefano
Passerini, Andrea
Machine Learning
Artificial Intelligence
As Graph Neural Networks (GNNs) become more pervasive, it becomes paramount to build reliable tools for explaining their predictions. A core desideratum is that explanations are \textit{faithful}, \ie that they portray an accurate picture of the GNN's reasoning process. However, a number of different faithfulness metrics exist, begging the question of what is faithfulness exactly and how to achieve it. We make three key contributions. We begin by showing that \textit{existing metrics are not interchangeable} -- \ie explanations attaining high faithfulness according to one metric may be unfaithful according to others -- and can systematically ignore important properties of explanations. We proceed to show that, surprisingly, \textit{optimizing for faithfulness is not always a sensible design goal}. Specifically, we prove that for injective regular GNN architectures, perfectly faithful explanations are completely uninformative. This does not apply to modular GNNs, such as self-explainable and domain-invariant architectures, prompting us to study the relationship between architectural choices and faithfulness. Finally, we show that \textit{faithfulness is tightly linked to out-of-distribution generalization}, in that simply ensuring that a GNN can correctly recognize the domain-invariant subgraph, as prescribed by the literature, does not guarantee that it is invariant unless this subgraph is also faithful.The code is publicly available on GitHub
title Reconsidering Faithfulness in Regular, Self-Explainable and Domain Invariant GNNs
topic Machine Learning
Artificial Intelligence
url https://arxiv.org/abs/2406.15156