mlsanityData to 14 Oct 2026

Exploring specialization and sensitivity of convolutional neural networks in the context of simultaneous image augmentations

arXiv 2503.03283 ↗First posted 5 Mar 2025stat.ML cs.AI cs.LGcited by 2

Not on the map. Why?

Abstract (arXiv)

Drawing parallels with the way biological networks are studied, we adapt the treatment--control paradigm to explainable artificial intelligence research and enrich it through multi-parametric input alterations. In this study, we propose a framework for investigating the internal inference impacted by input data augmentations. The internal changes in network operation are reflected in activation changes measured by variance, which can be decomposed into components related to each augmentation, employing Sobol indices and Shapley values. These quantities enable one to visualize sensitivity to different variables and use them for guided masking of activations. In addition, we introduce a way of single-class sensitivity analysis where the candidates are filtered according to their matching to prediction bias generated by targeted damaging of the activations. Relying on the observed parallels, we assume that the developed framework can potentially be transferred to studying biological neural networks in complex environments.

Key facts. Cited by 2 papers in the map (data through 14 Oct 2026). Builds on 33 of its 80 references that are in the map. First posted on arXiv 5 Mar 2025.

Builds on

33 of 80 references.

References · 33

  1. 2015Deep Residual Learning for Image Recognitioncited by 52,096
  2. 2020An Image is Worth 16x16 Words: Transformers for Image Recognition at Scalecited by 30,186
  3. 2019PyTorch: An Imperative Style, High-Performance Deep Learning Librarycited by 20,305
  4. 2014Very Deep Convolutional Networks for Large-Scale Image Recognitioncited by 15,364
  5. 2015Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networkscited by 9,639
  6. 2017A Unified Approach to Interpreting Model Predictionscited by 6,555
  7. 2016"Why Should I Trust You?": Explaining the Predictions of Any Classifiercited by 5,070
  8. 2014Neural Machine Translation by Jointly Learning to Align and Translatecited by 4,548
  9. 2014Learning Phrase Representations using RNN Encoder-Decoder for Statistical Machine Translationcited by 4,057
  10. 2019Benchmarking Neural Network Robustness to Common Corruptions and Perturbationscited by 2,877
  11. 2013Visualizing and Understanding Convolutional Networkscited by 2,290
  12. 2017Learning Important Features Through Propagating Activation Differencescited by 1,458
21 more
  1. 2019Natural Adversarial Examplescited by 1,408
  2. 2017Interpretability Beyond Feature Attribution: Quantitative Testing with Concept Activation Vectors (TCAV)cited by 1,319
  3. 2022LLM.int8(): 8-bit Matrix Multiplication for Transformers at Scalecited by 906
  4. 2018RISE: Randomized Input Sampling for Explanation of Black-box Modelscited by 838
  5. 2017Network Dissection: Quantifying Interpretability of Deep Visual Representationscited by 708
  6. 2022MaxViT: Multi-Axis Vision Transformercited by 476
  7. 2021VATT: Transformers for Multimodal Self-Supervised Learning from Raw Video, Audio and Textcited by 429
  8. 2015Evaluating the visualization of what a Deep Neural Network has learnedcited by 419
  9. 2014Understanding Deep Image Representations by Inverting Themcited by 403
  10. 2017Variational Dropout Sparsifies Deep Neural Networkscited by 265
  11. 2017Recent Trends in Deep Learning Based Natural Language Processingcited by 263
  12. 2018Generalisation in humans and deep neural networkscited by 237
  13. 2015Inverting Visual Representations with Convolutional Networkscited by 170
  14. 2018Benchmarking Neural Network Robustness to Common Corruptions and Surface Variationscited by 99
  15. 2020Principles and Practice of Explainable Machine Learningcited by 99
  16. 2020Convolutional Neural Networks as a Model of the Visual System: Past, Present, and Futurecited by 83
  17. 2017Tracking The Untrackable: Learning To Track Multiple Cues with Long-Term Dependenciescited by 76
  18. 2021Look at the Variance! Efficient Black-box Explanations with Sobol-based Sensitivity Analysiscited by 65
  19. 2020Speaker Recognition Based on Deep Learning: An Overviewcited by 59
  20. 2023Occlusion Sensitivity Analysis with Augmentation Subspace Perturbation in Deep Feature Spacecited by 6
  21. 2018Explaining Machine Learning Models using Entropic Variable Projectioncited by 5

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 1; 2026: 1.

2 papers.

  1. 2025Application of Sensitivity Analysis Methods for Studying Neural Network Modelscited by 0
  2. 2026Entropy and Attention Dynamics in Small Language Models: A Trace-Level Structural Analysis on the TruthfulQA Benchmarkcited by 0