Discriminative Attribution from Counterfactuals
Not on the map. Why?
Abstract (arXiv)
We present a method for neural network interpretability by combining feature attribution with counterfactual explanations to generate attribution maps that highlight the most discriminative features between pairs of classes. We show that this method can be used to quantitatively evaluate the performance of feature attribution methods in an objective manner, thus preventing potential observer bias. We evaluate the proposed method on three diverse datasets, including a challenging artificial dataset and real-world biological data. We show quantitatively and qualitatively that the highlighted features are substantially more discriminative than those extracted using conventional attribution methods and argue that this type of explanation is better suited for understanding fine grained class differences as learned by a deep neural network.
Key facts. Cited by 2 papers in the map (data through 7 Oct 2026). Builds on 34 of its 43 references that are in the map. First posted on arXiv 28 Sep 2021.
Builds on
34 of 43 references.
References · 34
- 2015Deep Residual Learning for Image Recognitioncited by 51,809
- 2014Adam: A Method for Stochastic Optimizationcited by 36,960
- 2016Grad-CAM: Visual Explanations from Deep Networks via Gradient-based Localizationcited by 6,690
- 2017A Unified Approach to Interpreting Model Predictionscited by 6,488
- 2016Image-to-Image Translation with Conditional Adversarial Networkscited by 4,510
- 2017Axiomatic Attribution for Deep Networkscited by 3,350
- 2016Improved Techniques for Training GANscited by 2,938
- 2013Deep Inside Convolutional Networks: Visualising Image Classification Models and Saliency Mapscited by 2,348
- 2013Visualizing and Understanding Convolutional Networkscited by 2,278
- 2016Understanding deep learning requires rethinking generalizationcited by 1,657
- 2017Learning Important Features Through Propagating Activation Differencescited by 1,454
- 2018Sanity Checks for Saliency Mapscited by 1,149
22 more
- 2017SmoothGrad: removing noise by adding noisecited by 1,097
- 2017Counterfactual Explanations without Opening the Black Box: Automated Decisions and the GDPRcited by 1,086
- 2016Least Squares Generative Adversarial Networkscited by 1,000
- 2014Striving for Simplicity: The All Convolutional Netcited by 983
- 2019Explaining Machine Learning Classifiers through Diverse Counterfactual Explanationscited by 612
- 2017Interpretable Explanations of Black Boxes by Meaningful Perturbationcited by 611
- 2018A Benchmark for Interpretability Methods in Deep Neural Networkscited by 467
- 2017Interpretation of Neural Networks is Fragilecited by 443
- 2015Evaluating the visualization of what a Deep Neural Network has learnedcited by 416
- 2017The (Un)reliability of saliency methodscited by 331
- 2019Counterfactual Visual Explanationscited by 312
- 2018On the Robustness of Interpretability Methodscited by 310
- 2016Not Just a Black Box: Learning Important Features Through Propagating Activation Differencescited by 295
- 2017Real Time Image Saliency for Black Box Classifierscited by 246
- 2016Model-Agnostic Interpretability of Machine Learningcited by 219
- 2017Visualizing Deep Neural Network Decisions: Prediction Difference Analysiscited by 166
- 2016Mode Regularized Generative Adversarial Networkscited by 123
- 2017Learning how to explain neural networks: PatternNet and PatternAttributioncited by 84
- 2020SCOUT: Self-aware Discriminant Counterfactual Explanationscited by 52
- 2019Generative Counterfactual Introspection for Explainable Deep Learningcited by 36
- 2020Scientific Discovery by Generating Counterfactuals using Image Translationcited by 11
- 2020Plausible Counterfactuals: Auditing Deep Learning Classifiers with Realistic Adversarial Examplescited by 9
Built on by
Citing papers per year: 2023: 1; 2024: 0; 2025: 0; 2026: 1.
2 papers.