mlsanityData to 7 Oct 2026

A Game-Theoretic Taxonomy of Visual Concepts in DNNs

arXiv 2106.10938 ↗First posted 21 Jun 2021cs.LG cs.CVcited by 16

Not on the map. Why?

Abstract (arXiv)

In this paper, we rethink how a DNN encodes visual concepts of different complexities from a new perspective, i.e. the game-theoretic multi-order interactions between pixels in an image. Beyond the categorical taxonomy of objects and the cognitive taxonomy of textures and shapes, we provide a new taxonomy of visual concepts, which helps us interpret the encoding of shapes and textures, in terms of concept complexities. In this way, based on multi-order interactions, we find three distinctive signal-processing behaviors of DNNs encoding textures. Besides, we also discover the flexibility for a DNN to encode shapes is lower than the flexibility of encoding textures. Furthermore, we analyze how DNNs encode outlier samples, and explore the impacts of network architectures on interactions. Additionally, we clarify the crucial role of the multi-order interactions in real-world applications. The code will be released when the paper is accepted.

Key facts. Cited by 16 papers in the map (data through 7 Oct 2026). Builds on 34 of its 45 references that are in the map. First posted on arXiv 21 Jun 2021.

Builds on

34 of 45 references.

References · 34

  1. 2015Deep Residual Learning for Image Recognitioncited by 51,809
  2. 2014ImageNet Large Scale Visual Recognition Challengecited by 8,441
  3. 2016Grad-CAM: Visual Explanations from Deep Networks via Gradient-based Localizationcited by 6,690
  4. 2017A Unified Approach to Interpreting Model Predictionscited by 6,488
  5. 2016"Why Should I Trust You?": Explaining the Predictions of Any Classifiercited by 5,040
  6. 2015Learning Deep Features for Discriminative Localizationcited by 2,430
  7. 2013Deep Inside Convolutional Networks: Visualising Image Classification Models and Saliency Mapscited by 2,348
  8. 2013Visualizing and Understanding Convolutional Networkscited by 2,278
  9. 2017Improved Regularization of Convolutional Neural Networks with Cutoutcited by 1,623
  10. 2017Interpretability Beyond Feature Attribution: Quantitative Testing with Concept Activation Vectors (TCAV)cited by 1,309
  11. 2017Grad-CAM++: Improved Visual Explanations for Deep Convolutional Networkscited by 936
  12. 2016Pixel Recurrent Neural Networkscited by 814
22 more
  1. 2017Network Dissection: Quantifying Interpretability of Deep Visual Representationscited by 703
  2. 2017Interpretable Explanations of Black Boxes by Meaningful Perturbationcited by 611
  3. 2018A Benchmark for Interpretability Methods in Deep Neural Networkscited by 467
  4. 2014Understanding Deep Image Representations by Inverting Themcited by 399
  5. 2015Understanding Neural Networks Through Deep Visualizationcited by 332
  6. 2018Explanations based on the Missing: Towards Contrastive Explanations with Pertinent Negativescited by 261
  7. 2019Explanations can be manipulated and geometry is to blamecited by 215
  8. 2016Synthesizing the preferred inputs for neurons in neural networks via deep generator networkscited by 181
  9. 2015Inverting Visual Representations with Convolutional Networkscited by 168
  10. 2017Interpretable Convolutional Neural Networkscited by 161
  11. 2017Visualizing and Understanding Atari Agentscited by 141
  12. 2018Net2Vec: Quantifying and Explaining how Concepts are Encoded by Filters in Deep Neural Networkscited by 128
  13. 2019XRAI: Better Attributions Through Regionscited by 116
  14. 2019Explaining Deep Neural Networks with a Polynomial Time Algorithm for Shapley Values Approximationcited by 115
  15. 2020Restricting the Flow: Information Bottlenecks for Attributioncited by 110
  16. 2017Learning how to explain neural networks: PatternNet and PatternAttributioncited by 84
  17. 2020Shape-Texture Debiased Neural Network Trainingcited by 76
  18. 2021Shape or Texture: Understanding Discriminative Features in CNNscited by 52
  19. 2019ML-LOO: Detecting Adversarial Examples with Feature Attributioncited by 48
  20. 2020Interpreting and Boosting Dropout from a Game-Theoretic Viewcited by 34
  21. 2020A Singular Value Perspective on Model Robustnesscited by 7
  22. 2020Interpreting Multivariate Shapley Interactions in DNNscited by 4

Built on by

Cited by year

Citing papers per year: 2021: 4; 2022: 2; 2023: 6; 2024: 2; 2025: 2.

16 papers.

Sort:

  1. 2022MogaNet: Multi-order Gated Aggregation Networkcited by 59
  2. 2021Discovering and Explaining the Representation Bottleneck of DNNscited by 42
  3. 2021Defining and Quantifying the Emergence of Sparse Concepts in DNNscited by 36
  4. 2023Does a Neural Network Really Encode Symbolic Concepts?cited by 27
  5. 2023Explaining Generalization Power of a DNN Using Interactive Conceptscited by 25
  6. 2021A Unified Game-Theoretic Interpretation of Adversarial Robustnesscited by 17
  7. 2024Towards the Dynamics of a DNN Learning Symbolic Interactionscited by 13
  8. 2023Can the Inference Logic of Large Language Models be Disentangled into Symbolic Concepts?cited by 11
  9. 2023Technical Note: Defining and Quantifying AND-OR Interactions for Faithful and Concise Explanation of DNNscited by 10
  10. 2023Where We Have Arrived in Proving the Emergence of Sparse Symbolic Concepts in AI Modelscited by 8
  11. 2021Visualizing the Emergence of Intermediate Visual Patterns in DNNscited by 7
  12. 2024Identifying Important Group of Pixels using Interactionscited by 5
4 more
  1. 2023Understanding Data Augmentation from a Robustness Perspectivecited by 3
  2. 2025A Survey on Interpretability in Visual Recognitioncited by 2
  3. 2022Game-Theoretic Understanding of Misclassificationcited by 1
  4. 2025SpaRTAN: Spatial Reinforcement Token-based Aggregation Network for Visual Recognitioncited by 0