mlsanityData to 7 Oct 2026

GradDiv: Adversarial Robustness of Randomized Neural Networks via Gradient Diversity Regularization

arXiv 2107.02425 ↗First posted 6 Jul 2021cs.LGcited by 34

Not on the map. Why?

Abstract (arXiv)

Deep learning is vulnerable to adversarial examples. Many defenses based on randomized neural networks have been proposed to solve the problem, but fail to achieve robustness against attacks using proxy gradients such as the Expectation over Transformation (EOT) attack. We investigate the effect of the adversarial attacks using proxy gradients on randomized neural networks and demonstrate that it highly relies on the directional distribution of the loss gradients of the randomized neural network. We show in particular that proxy gradients are less effective when the gradients are more scattered. To this end, we propose Gradient Diversity (GradDiv) regularizations that minimize the concentration of the gradients to build a robust randomized neural network. Our experiments on MNIST, CIFAR10, and STL10 show that our proposed GradDiv regularizations improve the adversarial robustness of randomized neural networks against a variety of state-of-the-art attack methods. Moreover, our method efficiently reduces the transferability among sample models of randomized neural networks.

Key facts. Cited by 34 papers in the map (data through 7 Oct 2026). Builds on 28 of its 36 references that are in the map. First posted on arXiv 6 Jul 2021.

Builds on

28 of 36 references.

References · 28

  1. 2014Explaining and Harnessing Adversarial Examplescited by 7,287
  2. 2017Towards Deep Learning Models Resistant to Adversarial Attackscited by 6,326
  3. 2013Intriguing properties of neural networkscited by 5,114
  4. 2016Towards Evaluating the Robustness of Neural Networkscited by 2,976
  5. 2019Theoretically Principled Trade-off between Robustness and Accuracycited by 1,510
  6. 2020Reliable evaluation of adversarial robustness with an ensemble of diverse parameter-free attackscited by 1,442
  7. 2018Obfuscated Gradients Give a False Sense of Security: Circumventing Defenses to Adversarial Examplescited by 1,259
  8. 2016Practical Black-Box Attacks against Machine Learningcited by 893
  9. 2019Square Attack: a query-efficient black-box adversarial attack via random searchcited by 682
  10. 2020On Adaptive Attacks to Adversarial Example Defensescited by 547
  11. 2016Delving into Transferable Adversarial Examples and Black-box Attackscited by 530
  12. 2017Countering Adversarial Images using Input Transformationscited by 517
16 more
  1. 2016Transferability in Machine Learning: from Phenomena to Black-Box Attacks using Adversarial Samplescited by 439
  2. 2017Mitigating Adversarial Effects Through Randomizationcited by 382
  3. 2019Minimally distorted Adversarial Examples with a Fast Adaptive Boundary Attackcited by 333
  4. 2020Uncovering the Limits of Adversarial Training against Norm-Bounded Adversarial Examplescited by 275
  5. 2019Improving Adversarial Robustness via Promoting Ensemble Diversitycited by 213
  6. 2018Stochastic Activation Pruning for Robust Adversarial Defensecited by 171
  7. 2017Towards Robust Neural Networks via Random Self-ensemblecited by 158
  8. 2017The Space of Transferable Adversarial Examplescited by 156
  9. 2018Parametric Noise Injection: Trainable Randomness to Improve Deep Neural Network Robustness against Adversarial Attackcited by 123
  10. 2020DVERGE: Diversifying Vulnerabilities for Enhanced Robust Generation of Ensemblescited by 73
  11. 2019Improving Adversarial Robustness of Ensembles with Diversity Trainingcited by 69
  12. 2018Adv-BNN: Improved Adversarial Defense through Robust Bayesian Neural Networkcited by 68
  13. 2019Mixup Inference: Better Exploiting Mixup to Defend Adversarial Attackscited by 54
  14. 2019Accurate, reliable and fast robustness evaluationcited by 43
  15. 2019Comment on "Adv-BNN: Improved Adversarial Defense through Robust Bayesian Neural Network"cited by 12
  16. 2020Weight-Covariance Alignment for Adversarially Robust Neural Networkscited by 9

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2021: 1; 2022: 5; 2023: 8; 2024: 10; 2025: 7; 2026: 3.

34 papers.

Sort:

By area:

  1. 2024Estimating Epistemic and Aleatoric Uncertainty with a Single Modelcited by 52
  2. 2023On the role of Model Uncertainties in Bayesian Optimizationcited by 7
  3. 2024Boosting Model Resilience via Implicit Adversarial Data Augmentationcited by 7
  4. 2023Survey on Leveraging Uncertainty Estimation Towards Trustworthy Deep Neural Networks: The Case of Reject Option and Post-training Processingcited by 6
  5. 2023Data Optimization in Deep Learning: A Surveycited by 6
  6. 2023Adversarial Robust Memory-Based Continual Learnercited by 5
  7. 2025Measuring Diversity in Synthetic Datasetscited by 5
  8. 2024Inferring the Langevin Equation with Uncertainty via Bayesian Neural Networkscited by 4
  9. 2023LMD: Light-weight Prediction Quality Estimation for Object Detection in Lidar Point Cloudscited by 3
  10. 2023NEO-KD: Knowledge-Distillation-Based Adversarial Training for Robust Multi-Exit Neural Networkscited by 3
  11. 2024Adversarially Robust Industrial Anomaly Detection Through Diffusion Modelcited by 3
  12. 2024Lightning UQ Box: A Comprehensive Framework for Uncertainty Quantification in Deep Learningcited by 3
22 more
  1. 2024MIRACLE3D: Memory-efficient Integrated Robust Approach for Continual Learning on Point Clouds via Shape Model Constructioncited by 3
  2. 2022Attacking Adversarial Defences by Smoothing the Loss Landscapecited by 2
  3. 2024Improving Adversarial Robustness via Decoupled Visual Representation Maskingcited by 2
  4. 2025Towards Trustworthy Vital Sign Forecasting: Leveraging Uncertainty for Prediction Intervalscited by 2
  5. 2022How Sampling Impacts the Robustness of Stochastic Neural Networkscited by 1
  6. 2022Gradient Obfuscation Checklist Test Gives a False Sense of Securitycited by 1
  7. 2022BayesNetCNN: incorporating uncertainty in neural networks for image-based classification taskscited by 1
  8. 2022Ares: A System-Oriented Wargame Framework for Adversarial MLcited by 1
  9. 2023Uncertainty Aware Neural Network from Similarity and Sensitivitycited by 1
  10. 2024BayesNAM: Leveraging Inconsistency for Reliable Explanationscited by 1
  11. 2025Bayesian Optimization with Preference Exploration using a Monotonic Neural Network Ensemblecited by 1
  12. 2026Provably Robust Bayesian Counterfactual Explanations under Model Changescited by 1
  13. 2026Spatio-Temporal Gaussian Process for Building Terrain-Incorporating Wind Power Curvescited by 1
  14. 2021Improving the Behaviour of Vision Transformers with Token-consistent Stochastic Layerscited by 0
  15. 2023On Continuity of Robust and Accurate Classifierscited by 0
  16. 2024Learning from Uncertain Data: From Possible Worlds to Possible Modelscited by 0
  17. 2024Comparative Evaluation of Applicability Domain Definition Methods for Regression Modelscited by 0
  18. 2025Feature Statistics with Uncertainty Help Adversarial Robustnesscited by 0
  19. 2025Partially stochastic deep learning with uncertainty quantification for model predictive heating controlcited by 0
  20. 2025Learning Optimal Prompt Ensemble for Multi-source Visual Prompt Transfercited by 0
  21. 2025The Sensitivity of Variational Bayesian Neural Network Performance to Hyperparameterscited by 0
  22. 2026Neural Aided Kalman Filtering for UAV State Estimation in Degraded Sensing Environmentscited by 0