Saved in:
Bibliographic Details
Main Authors: Simpson, Lachlan, Costanza, Federico, Millar, Kyle, Cheng, Adriel, Lim, Cheng-Chew, Chew, Hong Gunn
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2503.08240
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909534082039808
author Simpson, Lachlan
Costanza, Federico
Millar, Kyle
Cheng, Adriel
Lim, Cheng-Chew
Chew, Hong Gunn
author_facet Simpson, Lachlan
Costanza, Federico
Millar, Kyle
Cheng, Adriel
Lim, Cheng-Chew
Chew, Hong Gunn
contents Integrated gradients is prevalent within machine learning to address the black-box problem of neural networks. The explanations given by integrated gradients depend on a choice of base-point. The choice of base-point is not a priori obvious and can lead to drastically different explanations. There is a longstanding hypothesis that data lies on a low dimensional Riemannian manifold. The quality of explanations on a manifold can be measured by the extent to which an explanation for a point lies in its tangent space. In this work, we propose that the base-point should be chosen such that it maximises the tangential alignment of the explanation. We formalise the notion of tangential alignment and provide theoretical conditions under which a base-point choice will provide explanations lying in the tangent space. We demonstrate how to approximate the optimal base-point on several well-known image classification datasets. Furthermore, we compare the optimal base-point choice with common base-points and three gradient explainability models.
format Preprint
id arxiv_https___arxiv_org_abs_2503_08240
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Tangentially Aligned Integrated Gradients for User-Friendly Explanations
Simpson, Lachlan
Costanza, Federico
Millar, Kyle
Cheng, Adriel
Lim, Cheng-Chew
Chew, Hong Gunn
Machine Learning
Differential Geometry
Integrated gradients is prevalent within machine learning to address the black-box problem of neural networks. The explanations given by integrated gradients depend on a choice of base-point. The choice of base-point is not a priori obvious and can lead to drastically different explanations. There is a longstanding hypothesis that data lies on a low dimensional Riemannian manifold. The quality of explanations on a manifold can be measured by the extent to which an explanation for a point lies in its tangent space. In this work, we propose that the base-point should be chosen such that it maximises the tangential alignment of the explanation. We formalise the notion of tangential alignment and provide theoretical conditions under which a base-point choice will provide explanations lying in the tangent space. We demonstrate how to approximate the optimal base-point on several well-known image classification datasets. Furthermore, we compare the optimal base-point choice with common base-points and three gradient explainability models.
title Tangentially Aligned Integrated Gradients for User-Friendly Explanations
topic Machine Learning
Differential Geometry
url https://arxiv.org/abs/2503.08240