The Manifold Hypothesis for Gradient-Based Explanations. (arXiv:2206.07387v1 [cs.LG])

Web: http://arxiv.org/abs/2206.07387

When do gradient-based explanation algorithms provide meaningful
explanations? We propose a necessary criterion: their feature attributions need
to be aligned with the tangent space of the data manifold. To provide evidence
for this hypothesis, we introduce a framework based on variational autoencoders
that allows to estimate and generate image manifolds. Through experiments
across a range of different datasets -- MNIST, EMNIST, CIFAR10, X-ray pneumonia
and Diabetic Retinopathy detection -- we demonstrate that the more a feature
attribution is aligned with …

arxiv gradient hypothesis lg manifold