mlsanityData to 14 Oct 2026

Interpretable Neural Networks with Frank-Wolfe: Sparse Relevance Maps and Relevance Orderings

arXiv 2110.08105 ↗First posted 15 Oct 2021cs.LG math.OCcited by 5

Not on the map. Why?

Abstract (arXiv)

We study the effects of constrained optimization formulations and Frank-Wolfe algorithms for obtaining interpretable neural network predictions. Reformulating the Rate-Distortion Explanations (RDE) method for relevance attribution as a constrained optimization problem provides precise control over the sparsity of relevance maps. This enables a novel multi-rate as well as a relevance-ordering variant of RDE that both empirically outperform standard RDE and other baseline methods in a well-established comparison test. We showcase several deterministic and stochastic variants of the Frank-Wolfe algorithm and their effectiveness for RDE.

Key facts. Cited by 5 papers in the map (data through 14 Oct 2026). Builds on 30 of its 58 references that are in the map. First posted on arXiv 15 Oct 2021.

Builds on

30 of 58 references.

References · 30

  1. 2017Attention Is All You Needcited by 60,361
  2. 2014Very Deep Convolutional Networks for Large-Scale Image Recognitioncited by 15,364
  3. 2017A Unified Approach to Interpreting Model Predictionscited by 6,555
  4. 2016"Why Should I Trust You?": Explaining the Predictions of Any Classifiercited by 5,070
  5. 2014Learning Phrase Representations using RNN Encoder-Decoder for Statistical Machine Translationcited by 4,057
  6. 2013Deep Inside Convolutional Networks: Visualising Image Classification Models and Saliency Mapscited by 2,371
  7. 2017Towards A Rigorous Science of Interpretable Machine Learningcited by 1,416
  8. 2017SmoothGrad: removing noise by adding noisecited by 1,109
  9. 2014Striving for Simplicity: The All Convolutional Netcited by 990
  10. 2017Interpretable Explanations of Black Boxes by Meaningful Perturbationcited by 612
  11. 2018Explaining Explanations: An Overview of Interpretability of Machine Learningcited by 508
  12. 2015Evaluating the visualization of what a Deep Neural Network has learnedcited by 419
18 more
  1. 2017Deep Learning for Medical Image Analysiscited by 401
  2. 2017Methods for Interpreting and Understanding Deep Neural Networkscited by 345
  3. 2020A Survey on Neural Network Interpretabilitycited by 291
  4. 2019Feature relevance quantification in explainable AI: A causal problemcited by 167
  5. 2020On Interpretability of Artificial Neural Networks: A Surveycited by 88
  6. 2015On the Global Linear Convergence of Frank-Wolfe Optimization Variantscited by 77
  7. 2016Convergence Rate of Frank-Wolfe for Non-Convex Objectivescited by 63
  8. 2018Stochastic Conditional Gradient Methods: From Convex Minimization to Submodular Maximizationcited by 47
  9. 2016Stochastic Frank-Wolfe Methods for Nonconvex Optimizationcited by 41
  10. 2016Variance-Reduced and Projection-Free Stochastic Optimizationcited by 40
  11. 2019A Rate-Distortion Framework for Explaining Neural Network Decisionscited by 24
  12. 2020Deep Neural Network Training with Frank-Wolfecited by 23
  13. 2021Complexity of Linear Minimization and Projection on Some Setscited by 20
  14. 2020Stochastic Frank-Wolfe for Constrained Finite-Sum Minimizationcited by 14
  15. 2018Deep Frank-Wolfe For Neural Network Optimizationcited by 11
  16. 2016Lazifying Conditional Gradient Algorithmscited by 8
  17. 2019Stochastic Conditional Gradient++cited by 8
  18. 2021FrankWolfe.jl: a high-performance and flexible toolbox for Frank-Wolfe algorithms and Conditional Gradientscited by 8

Built on by

Cited by year

Citing papers per year: 2022: 4; 2023: 0; 2024: 0; 2025: 1.

5 papers.

  1. 2022Spatial-temporal Concept based Explanation of 3D ConvNetscited by 12
  2. 2022Interpretability Guarantees with Merlin-Arthur Classifierscited by 9
  3. 2022Training Characteristic Functions with Reinforcement Learning: XAI-methods play Connect Fourcited by 3
  4. 2022SAIF: Sparse Adversarial and Imperceptible Attack Frameworkcited by 2
  5. 2025Implicit Bias in Matrix Factorization and its Explicit Realization in a New Architecturecited by 1