mlsanityData to 14 Oct 2026

Gradient-Based Interpretability Methods and Binarized Neural Networks

arXiv 2106.12569 ↗First posted 23 Jun 2021cs.CV cs.AI cs.LGcited by 1

Not on the map. Why?

Abstract (arXiv)

Binarized Neural Networks (BNNs) have the potential to revolutionize the way that deep learning is carried out in edge computing platforms. However, the effectiveness of interpretability methods on these networks has not been assessed. In this paper, we compare the performance of several widely used saliency map-based interpretabilty techniques (Gradient, SmoothGrad and GradCAM), when applied to Binarized or Full Precision Neural Networks (FPNNs). We found that the basic Gradient method produces very similar-looking maps for both types of network. However, SmoothGrad produces significantly noisier maps for BNNs. GradCAM also produces saliency maps which differ between network types, with some of the BNNs having seemingly nonsensical explanations. We comment on possible reasons for these differences in explanations and present it as an example of why interpretability techniques should be tested on a wider range of network types.

Key facts. Cited by 1 paper in the map (data through 14 Oct 2026). Builds on 16 of its 23 references that are in the map. First posted on arXiv 23 Jun 2021.

Builds on

16 of 23 references.

References · 16

  1. 2015Deep Residual Learning for Image Recognitioncited by 52,096
  2. 2014Very Deep Convolutional Networks for Large-Scale Image Recognitioncited by 15,364
  3. 2016Grad-CAM: Visual Explanations from Deep Networks via Gradient-based Localizationcited by 6,733
  4. 2019EfficientNet: Rethinking Model Scaling for Convolutional Neural Networkscited by 6,442
  5. 2018MobileNetV2: Inverted Residuals and Linear Bottleneckscited by 4,795
  6. 2017MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applicationscited by 3,845
  7. 2013Deep Inside Convolutional Networks: Visualising Image Classification Models and Saliency Mapscited by 2,371
  8. 2018Sanity Checks for Saliency Mapscited by 1,156
  9. 2017SmoothGrad: removing noise by adding noisecited by 1,109
  10. 2017Towards Accurate Binary Convolutional Neural Networkcited by 149
  11. 2020Training Binary Neural Networks with Real-to-Binary Convolutionscited by 112
  12. 2019Defensive Quantization: When Efficiency Meets Robustnesscited by 95
4 more
  1. 2020MeliusNet: Can Binary Neural Networks Achieve MobileNet-level Accuracy?cited by 32
  2. 2017Attacking Binarized Neural Networkscited by 29
  3. 2019Back to Simplicity: How to Train Accurate BNNs from Scratch?cited by 24
  4. 2020Larq Compute Engine: Design, Benchmark, and Deploy State-of-the-Art Binarized Neural Networkscited by 20

Built on by

Cited by year

Citing papers per year: 2025: 1.

1 paper.

  1. 2025CRITS: Convolutional Rectifier for Interpretable Time Series Classificationcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2025Understanding Generalization, Robustness, and Interpretability in Low-Capacity Neural Networks
  2. 2022Join the High Accuracy Club on ImageNet with A Binary Neural Network Ticket
  3. 2025A Comprehensive Survey on Self-Interpretable Neural Networks
  4. 2024Saliency Assisted Quantization for Neural Networks
  5. 2022LAB: Learnable Activation Binarizer for Binary Neural Networks
  6. 2023Input Layer Binarization with Bit-Plane Encoding
  7. 2024Investigating the Effect of Network Pruning on Performance and Interpretability
  8. 2023Rethinking interpretation: Input-agnostic saliency mapping of deep visual classifiers