Intrinsic Biologically Plausible Adversarial Robustness
Not on the map. Why?
Abstract (arXiv)
Artificial Neural Networks (ANNs) trained with Backpropagation (BP) excel in different daily tasks but have a dangerous vulnerability: inputs with small targeted perturbations, also known as adversarial samples, can drastically disrupt their performance. Adversarial training, a technique in which the training dataset is augmented with exemplary adversarial samples, is proven to mitigate this problem but comes at a high computational cost. In contrast to ANNs, humans are not susceptible to misclassifying these same adversarial samples. Thus, one can postulate that biologically-plausible trained ANNs might be more robust against adversarial attacks. In this work, we chose the biologically-plausible learning algorithm Present the Error to Perturb the Input To modulate Activity (PEPITA) as a case study and investigated this question through a comparative analysis with BP-trained ANNs on various computer vision tasks. We observe that PEPITA has a higher intrinsic adversarial robustness and, when adversarially trained, also has a more favorable natural-vs-adversarial performance trade-off. In particular, for the same natural accuracies on the MNIST task, PEPITA's adversarial accuracies decrease on average only by 0.26% while BP's decrease by 8.05%.
Key facts. Cited by 1 paper in the map (data through 7 Oct 2026). Builds on 39 of its 53 references that are in the map. First posted on arXiv 29 Sep 2023.
Builds on
39 of 53 references.
References · 39
- 2014Explaining and Harnessing Adversarial Examplescited by 7,287
- 2017Towards Deep Learning Models Resistant to Adversarial Attackscited by 6,326
- 2013Intriguing properties of neural networkscited by 5,114
- 2017Fashion-MNIST: a Novel Image Dataset for Benchmarking Machine Learning Algorithmscited by 4,283
- 2015Delving Deep into Rectifiers: Surpassing Human-Level Performance on ImageNet Classificationcited by 3,590
- 2016Adversarial examples in the physical worldcited by 1,762
- 2019Theoretically Principled Trade-off between Robustness and Accuracycited by 1,510
- 2018Robustness May Be at Odds with Accuracycited by 873
- 2020Fast is better than free: Revisiting adversarial trainingcited by 723
- 2018Threat of Adversarial Attacks on Deep Learning in Computer Vision: A Surveycited by 459
- 2019Unlabeled Data Improves Adversarial Robustnesscited by 415
- 2022The Forward-Forward Algorithm: Some Preliminary Investigationscited by 262
27 more
- 2016Equilibrium Propagation: Bridging the Gap Between Energy-Based Models and Backpropagationcited by 251
- 2019Adversarial Policies: Attacking Deep Reinforcement Learningcited by 209
- 2023Better Diffusion Models Further Improve Adversarial Trainingcited by 207
- 2017Robust Deep Reinforcement Learning with Adversarial Attackscited by 152
- 2014Difference Target Propagationcited by 144
- 2018Dendritic cortical microcircuits approximate the backpropagation algorithmcited by 96
- 2019Deep Learning without Weight Transportcited by 87
- 2020Understanding Catastrophic Overfitting in Single-step Adversarial Trainingcited by 75
- 2017Robust Physical-World Attacks on Deep Learning Modelscited by 70
- 2019advertorch v0.1: An Adversarial Robustness Toolbox based on PyTorchcited by 69
- 2022Error-driven Input Modulation: Solving the Credit Assignment Problem without a Backward Passcited by 63
- 2019Efficient Adversarial Training with Transferable Adversarial Examplescited by 54
- 2022Efficient and Effective Augmentation Strategy for Adversarial Trainingcited by 43
- 2023Data Augmentation Alone Can Improve Adversarial Trainingcited by 43
- 2021WaveGuard: Understanding and Mitigating Audio Adversarial Examplescited by 39
- 2020Online Spatio-Temporal Learning in Deep Neural Networkscited by 32
- 2018Humans can decipher adversarial imagescited by 31
- 2021Credit Assignment in Neural Networks through Deep Feedback Controlcited by 28
- 2022Explicit Tradeoffs between Adversarial and Natural Distributional Robustnesscited by 20
- 2022Robustness in deep learning: The good (width), the bad (depth), and the ugly (initialization)cited by 18
- 2021ZeroGrad : Mitigating and Explaining Catastrophic Overfitting in FGSM Adversarial Trainingcited by 14
- 2023Forward Learning with Top-Down Feedback: Empirical and Analytical Characterizationcited by 13
- 2021Understanding Catastrophic Overfitting in Adversarial Trainingcited by 11
- 2022Efficient Adversarial Training With Data Pruningcited by 7
- 2019On the Adversarial Robustness of Neural Networks without Weight Transportcited by 5
- 2020Adversarial attacks on audio source separationcited by 5
- 2023A Review of Adversarial Attacks in Computer Visioncited by 5
Built on by
Citing papers per year: 2023: 1.
1 paper.