mlsanityData to 7 Oct 2026

Measuring Representational Robustness of Neural Networks Through Shared Invariances

arXiv 2206.11939 ↗First posted 23 Jun 2022cs.LG cs.AIcited by 11

Not on the map. Why?

Abstract (arXiv)

A major challenge in studying robustness in deep learning is defining the set of ``meaningless'' perturbations to which a given Neural Network (NN) should be invariant. Most work on robustness implicitly uses a human as the reference model to define such perturbations. Our work offers a new view on robustness by using another reference NN to define the set of perturbations a given NN should be invariant to, thus generalizing the reliance on a reference ``human NN'' to any NN. This makes measuring robustness equivalent to measuring the extent to which two NNs share invariances, for which we propose a measure called STIR. STIR re-purposes existing representation similarity measures to make them suitable for measuring shared invariances. Using our measure, we are able to gain insights into how shared invariances vary with changes in weight initialization, architecture, loss functions, and training dataset. Our implementation is available at: urlhttps://github.com/nvedant07/STIR.

Key facts. Cited by 11 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 23 Jun 2022.

Builds on

34 of 45 references.

References · 34

  1. 2015Deep Residual Learning for Image Recognitioncited by 51,809
  2. 2014Very Deep Convolutional Networks for Large-Scale Image Recognitioncited by 15,267
  3. 2018The Unreasonable Effectiveness of Deep Features as a Perceptual Metriccited by 11,029
  4. 2014Explaining and Harnessing Adversarial Examplescited by 7,287
  5. 2017Towards Deep Learning Models Resistant to Adversarial Attackscited by 6,326
  6. 2013Intriguing properties of neural networkscited by 5,114
  7. 2019Benchmarking Neural Network Robustness to Common Corruptions and Perturbationscited by 2,857
  8. 2019Similarity of Neural Network Representations Revisitedcited by 1,695
  9. 2019Theoretically Principled Trade-off between Robustness and Accuracycited by 1,510
  10. 2019Do ImageNet Classifiers Generalize to ImageNet?cited by 1,507
  11. 2020WILDS: A Benchmark of in-the-Wild Distribution Shiftscited by 1,366
  12. 2017Interpretability Beyond Feature Attribution: Quantitative Testing with Concept Activation Vectors (TCAV)cited by 1,309
22 more
  1. 2019Adversarial Examples Are Not Bugs, They Are Featurescited by 950
  2. 2018Robustness May Be at Odds with Accuracycited by 873
  3. 2015The Limitations of Deep Learning in Adversarial Settingscited by 784
  4. 2017Evasion Attacks against Machine Learning at Test Timecited by 686
  5. 2021Do Vision Transformers See Like Convolutional Neural Networks?cited by 671
  6. 2018Do Better ImageNet Models Transfer Better?cited by 560
  7. 2020Measuring Robustness to Natural Distribution Shifts in Image Classificationcited by 452
  8. 2014Understanding Deep Image Representations by Inverting Themcited by 399
  9. 2020What is being transferred in transfer learning?cited by 365
  10. 2018Insights on representational similarity in neural networks with canonical correlationcited by 318
  11. 2018Towards a Definition of Disentangled Representationscited by 294
  12. 2020Do Wide and Deep Networks Learn the Same Things? Uncovering How Neural Network Representations Vary with Width and Depthcited by 241
  13. 2018Generalisation in humans and deep neural networkscited by 237
  14. 2016Generating Images with Perceptual Similarity Metrics based on Deep Networkscited by 221
  15. 2015Convergent Learning: Do different neural networks learn the same representations?cited by 215
  16. 2021Revisiting Model Stitching to Compare Neural Representationscited by 203
  17. 2015Inverting Visual Representations with Convolutional Networkscited by 168
  18. 2018Towards Understanding Learning Representations: To What Extent Do Different Neural Networks Learn the Same Representationcited by 62
  19. 2021Grounding Representation Similarity with Statistical Testingcited by 28
  20. 2015Manitest: Are classifiers really invariant?cited by 27
  21. 2021On the Lack of Robust Interpretability of Neural Text Classifierscited by 17
  22. 2019Controversial stimuli: pitting neural networks against each other as models of human recognitioncited by 7

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2022: 1; 2023: 6; 2024: 2; 2025: 1; 2026: 1.

11 papers.

  1. 2023Getting aligned on representational alignmentcited by 142
  2. 2023Similarity of Neural Network Models: A Survey of Functional and Representational Measurescited by 128
  3. 2022Mechanistic Mode Connectivitycited by 50
  4. 2024Tracing Representation Progression: Analyzing and Enhancing Layer-Wise Similaritycited by 36
  5. 2023On Robustness of Finetuned Transformer-based NLP Modelscited by 8
  6. 2023Diffused Redundancy in Pre-trained Representationscited by 5
  7. 2023What Happens During Finetuning of Vision Transformers: An Invariance Based Investigationcited by 3
  8. 2024Understanding the Role of Invariance in Transfer Learningcited by 1
  9. 2023Perturbed examples reveal invariances shared by language modelscited by 0
  10. 2025Synthesizing Images on Perceptual Boundaries of ANNs for Uncovering Human Perceptual Variability on Facial Expressionscited by 0
  11. 2026TopoAlign: Topology-Aware Visual Representation Alignmentcited by 0