Robust Explainability: A tutorial on gradient-based attribution methods for deep neural networks
Authors:
- 28 June 2022
- journal article
- research article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Signal Processing Magazine
- Vol. 39 (4) , 73-84
Abstract
The rise in deep neural networks (DNNs) has led to increased interest in explaining their predictions. While many methods for this exist, there is currently no consensus on how to evaluate them. On the other hand, robustness is a popular topic for deep learning (DL) research; however, it has been hardly talked about in explainability until very recently.Keywords
All Related Versions
This publication has 18 references indexed in Scilit:
- The (Un)reliability of Saliency Methods2019
- Distorted Representation Space Characterization Through Backpropagated Gradients2019
- Interpretation of Neural Networks Is FragileProceedings of the AAAI Conference on Artificial Intelligence, 2019
- Stop explaining black box machine learning models for high stakes decisions and use interpretable models insteadNature Machine Intelligence, 2019
- Grad-CAM: Visual Explanations from Deep Networks via Gradient-Based Localization2017
- European Union Regulations on Algorithmic Decision Making and a “Right to Explanation”AI Magazine, 2017
- Explaining nonlinear classification decisions with deep Taylor decompositionPattern Recognition, 2017
- Evaluating the Visualization of What a Deep Neural Network Has LearnedIEEE Transactions on Neural Networks and Learning Systems, 2016
- On Pixel-Wise Explanations for Non-Linear Classifier Decisions by Layer-Wise Relevance PropagationPLOS ONE, 2015
- Visualizing and Understanding Convolutional Networks2014