mlsanityData to 7 Oct 2026

Stochastic Activation Pruning for Robust Adversarial Defense

arXiv 1803.01442 ↗First posted 5 Mar 2018cs.LG stat.MLcited by 171

Not on the map. Why?

Abstract (arXiv)

Neural networks are known to be vulnerable to adversarial examples. Carefully chosen perturbations to real images, while imperceptible to humans, induce misclassification and threaten the reliability of deep learning systems in the wild. To guard against adversarial examples, we take inspiration from game theory and cast the problem as a minimax zero-sum game between the adversary and the model. In general, for such games, the optimal strategy for both players requires a stochastic policy, also known as a mixed strategy. In this light, we propose Stochastic Activation Pruning (SAP), a mixed strategy for adversarial defense. SAP prunes a random subset of activations (preferentially pruning those with smaller magnitude) and scales up the survivors to compensate. We can apply SAP to pretrained networks, including adversarially trained models, without fine-tuning, providing robustness against adversarial examples. Experiments demonstrate that SAP confers robustness against attacks, increasing accuracy and preserving calibration.

Key facts. Cited by 171 papers in the map (data through 7 Oct 2026). First posted on arXiv 5 Mar 2018.

Builds on

No reference list yet.

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2021: 61; 2022: 42; 2023: 29; 2024: 20; 2025: 14; 2026: 5.

171 papers.

Sort:

By area:

  1. 2023A Simple and Effective Pruning Approach for Large Language Modelscited by 776
  2. 2021GAN Inversion: A Surveycited by 366
  3. 2023Certifying LLM Safety against Adversarial Promptingcited by 248
  4. 2022Extremely Simple Activation Shaping for Out-of-Distribution Detectioncited by 212
  5. 2021Adversarial purification with Score-based generative modelscited by 132
  6. 2022Guided Diffusion Model for Adversarial Purificationcited by 90
  7. 2021Improving Adversarial Robustness via Channel-wise Activation Suppressingcited by 76
  8. 2021Exploring Architectural Ingredients of Adversarially Robust Deep Neural Networkscited by 73
  9. 2021Lost in Pruning: The Effects of Pruning Neural Networks beyond Test Accuracycited by 62
  10. 2021Meta Gradient Adversarial Attackcited by 55
  11. 2021Adversarial Attacks are Reversible with Natural Supervisioncited by 47
  12. 2022Practical Evaluation of Adversarial Robustness via Adaptive Auto Attackcited by 44
38 more
  1. 2022A Survey of Robust Adversarial Training in Pattern Recognition: Fundamental, Theory, and Methodologiescited by 44
  2. 2022Sparsity Winning Twice: Better Robust Generalization from More Efficient Trainingcited by 41
  3. 2021Biometrics: Trust, but Verifycited by 39
  4. 2021Policy Smoothing for Provably Robust Reinforcement Learningcited by 39
  5. 2021Adversarial Robustness with Semi-Infinite Constrained Learningcited by 39
  6. 2021Inspect, Understand, Overcome: A Survey of Practical Methods for AI Safetycited by 35
  7. 2021GradDiv: Adversarial Robustness of Randomized Neural Networks via Gradient Diversity Regularizationcited by 34
  8. 2021Attacking Adversarial Attacks as A Defensecited by 31
  9. 2021Low Curvature Activations Reduce Overfitting in Adversarial Trainingcited by 29
  10. 2021A Review of Adversarial Attack and Defense for Classification Methodscited by 29
  11. 2022Towards Efficient Adversarial Training on Vision Transformerscited by 29
  12. 2023Integration of Large Language Models and Federated Learningcited by 28
  13. 2021Mixed Nash Equilibria in the Adversarial Examples Gamecited by 22
  14. 2021CIFS: Improving Adversarial Robustness of CNNs via Channel-wise Importance-based Feature Selectioncited by 21
  15. 2022On the Limitations of Stochastic Pre-processing Defensescited by 21
  16. 2022Defensive Patches for Robust Recognition in the Physical Worldcited by 20
  17. 2021Fighting Gradients with Gradients: Dynamic Defenses against Adversarial Attackscited by 19
  18. 2024Beyond Dropout: Robust Convolutional Neural Networks Based on Local Feature Maskingcited by 19
  19. 2021Center Smoothing: Certified Robustness for Networks with Structured Outputscited by 18
  20. 2021SoK: Anti-Facial Recognition Technologycited by 16
  21. 2022Improving Diversity with Adversarially Learned Transformations for Domain Generalizationcited by 16
  22. 2022PointDP: Diffusion-driven Purification against Adversarial Attacks on 3D Point Cloud Recognitioncited by 16
  23. 2023Feature Separation and Recalibration for Adversarial Robustnesscited by 16
  24. 2022Quarantine: Sparsity Can Uncover the Trojan Attack Trigger for Freecited by 15
  25. 2022Demystifying the Adversarial Robustness of Random Transformation Defensescited by 14
  26. 2025AutoAdvExBench: Benchmarking autonomous exploitation of adversarial example defensescited by 14
  27. 2021Local Competition and Stochasticity for Adversarial Robustness in Deep Learningcited by 12
  28. 2021ADC: Adversarial attacks against object Detection that evade Context consistency checkscited by 12
  29. 2022Rethinking Bayesian Learning for Data Analysis: The Art of Prior and Inference in Sparsity-Aware Modelingcited by 12
  30. 2022Understanding Adversarial Robustness Against On-manifold Adversarial Examplescited by 12
  31. 2023Improving Adversarial Robustness of Masked Autoencoders via Test-time Frequency-domain Promptingcited by 12
  32. 2021Adversarial Feature Augmentation and Normalization for Visual Recognitioncited by 11
  33. 2021On the Impact of Hard Adversarial Instances on Overfitting in Adversarial Trainingcited by 11
  34. 2021Input-Specific Robustness Certification for Randomized Smoothingcited by 11
  35. 2022Analysis and Extensions of Adversarial Training for Video Classificationcited by 11
  36. 2022An Evolutionary, Gradient-Free, Query-Efficient, Black-Box Algorithm for Generating Adversarial Instances in Deep Networkscited by 11
  37. 2021Resilient Machine Learning for Networked Cyber Physical Systems: A Survey for Machine Learning Security to Securing Machine Learning for CPScited by 10
  38. 2022Linearity Grafting: Relaxed Neuron Pruning Helps Certifiable Robustnesscited by 10

50 most cited of 171 All 171 (JSON)