mlsanityData to 7 Oct 2026

IDGI: A Framework to Eliminate Explanation Noise from Integrated Gradients

arXiv 2303.14242 ↗First posted 24 Mar 2023cs.CV cs.AI cs.LGcited by 16

Not on the map. Why?

Abstract (arXiv)

Integrated Gradients (IG) as well as its variants are well-known techniques for interpreting the decisions of deep neural networks. While IG-based approaches attain state-of-the-art performance, they often integrate noise into their explanation saliency maps, which reduce their interpretability. To minimize the noise, we examine the source of the noise analytically and propose a new approach to reduce the explanation noise based on our analytical findings. We propose the Important Direction Gradient Integration (IDGI) framework, which can be easily incorporated into any IG-based method that uses the Reimann Integration for integrated gradient computation. Extensive experiments with three IG-based methods show that IDGI improves them drastically on numerous interpretability metrics.

Key facts. Cited by 16 papers in the map (data through 7 Oct 2026). Builds on 33 of its 52 references that are in the map. First posted on arXiv 24 Mar 2023.

Builds on

33 of 52 references.

References · 33

  1. 2015Deep Residual Learning for Image Recognitioncited by 51,809
  2. 2014Very Deep Convolutional Networks for Large-Scale Image Recognitioncited by 15,267
  3. 2016Densely Connected Convolutional Networkscited by 6,691
  4. 2016Grad-CAM: Visual Explanations from Deep Networks via Gradient-based Localizationcited by 6,690
  5. 2017A Unified Approach to Interpreting Model Predictionscited by 6,488
  6. 2015Rethinking the Inception Architecture for Computer Visioncited by 5,451
  7. 2016"Why Should I Trust You?": Explaining the Predictions of Any Classifiercited by 5,040
  8. 2019nuScenes: A multimodal dataset for autonomous drivingcited by 4,485
  9. 2017MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applicationscited by 3,827
  10. 2017Axiomatic Attribution for Deep Networkscited by 3,350
  11. 2013Deep Inside Convolutional Networks: Visualising Image Classification Models and Saliency Mapscited by 2,348
  12. 2016Xception: Deep Learning with Depthwise Separable Convolutionscited by 2,320
21 more
  1. 2013Visualizing and Understanding Convolutional Networkscited by 2,278
  2. 2017Learning Important Features Through Propagating Activation Differencescited by 1,454
  3. 2018Sanity Checks for Saliency Mapscited by 1,149
  4. 2017SmoothGrad: removing noise by adding noisecited by 1,097
  5. 2014Striving for Simplicity: The All Convolutional Netcited by 983
  6. 2018RISE: Randomized Input Sampling for Explanation of Black-box Modelscited by 825
  7. 2016Multi-View 3D Object Detection Network for Autonomous Drivingcited by 721
  8. 2017Interpretable Explanations of Black Boxes by Meaningful Perturbationcited by 611
  9. 2016Conditional Image Synthesis With Auxiliary Classifier GANscited by 562
  10. 2017Interpretation of Neural Networks is Fragilecited by 443
  11. 2019The many Shapley values for model explanationcited by 341
  12. 2017Real Time Image Saliency for Black Box Classifierscited by 246
  13. 2015DeepDriving: Learning Affordance for Direct Perception in Autonomous Drivingcited by 229
  14. 2017Visualizing Deep Neural Network Decisions: Prediction Difference Analysiscited by 166
  15. 2021FastSHAP: Real-Time Shapley Value Estimationcited by 123
  16. 2019XRAI: Better Attributions Through Regionscited by 116
  17. 2021Guided Integrated Gradients: An Adaptive Path Method for Removing Noisecited by 102
  18. 2020Attribution in Scale and Spacecited by 61
  19. 2018Review of Visual Saliency Detection with Comprehensive Informationcited by 60
  20. 2019Visualizing Deep Networks by Optimizing with Integrated Gradientscited by 59
  21. 2020iGOS++: Integrated Gradient Optimized Saliency by Bilateral Perturbationscited by 19

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2023: 2; 2024: 8; 2025: 4; 2026: 2.

16 papers.

Sort:

  1. 2024Manifold Integrated Gradients: Riemannian Geometry for Feature Attributioncited by 21
  2. 2024Post-hoc Part-prototype Networkscited by 7
  3. 2025Tokenize Image Patches: Global Context Fusion for Effective Haze Removal in Large Imagescited by 7
  4. 2024Feature Attribution with Necessity and Sufficiency via Dual-stage Perturbation Test for Causal Explanationcited by 6
  5. 2023LICO: Explainable Models with Language-Image Consistencycited by 5
  6. 2024Supervised Fine-Tuning Achieve Rapid Task Adaption Via Alternating Attention Head Activation Patternscited by 5
  7. 2024QUCE: The Minimisation and Quantification of Path-Based Uncertainty for Generative Counterfactual Explanationscited by 4
  8. 2024AdaptGrad: Adaptive Sampling to Reduce Noisecited by 2
  9. 2025Time series saliency maps: explaining models across multiple domainscited by 2
  10. 2025Enhancing Interpretability for Vision Models via Shapley Value Optimizationcited by 2
  11. 2023SAfER: Layer-Level Sensitivity Assessment for Efficient and Robust Neural Network Inferencecited by 1
  12. 2024Riemann Sum Optimization for Accurate Integrated Gradients Computationcited by 1
4 more
  1. 2024Leveraging Local Structure for Improving Model Explanations: An Information Propagation Approachcited by 0
  2. 2025Enhancing Visual Feature Attribution via Weighted Integrated Gradientscited by 0
  3. 2026Training for Trustworthy Saliency Maps: Adversarial Training Meets Feature-Map Smoothingcited by 0
  4. 2026IG-Lens: Exact Additive Probability Attribution Across Transformer Layers via Telescoping Integrated Gradientscited by 0