mlsanityData to 14 Oct 2026

Noise as a Double-Edged Sword: Reinforcement Learning Exploits Randomized Defenses in Neural Networks

arXiv 2410.23870 ↗First posted 31 Oct 2024cs.CR cs.LGcited by 0

On the map: Deep reinforcement learning methods → Adversarial robustness and RL in finance

Abstract (arXiv)

This study investigates a counterintuitive phenomenon in adversarial machine learning: the potential for noise-based defenses to inadvertently aid evasion attacks in certain scenarios. While randomness is often employed as a defensive strategy against adversarial examples, our research reveals that this approach can sometimes backfire, particularly when facing adaptive attackers using reinforcement learning (RL). Our findings show that in specific cases, especially with visually noisy classes, the introduction of noise in the classifier's confidence values can be exploited by the RL attacker, leading to a significant increase in evasion success rates. In some instances, the noise-based defense scenario outperformed other strategies by up to 20% on a subset of classes. However, this effect was not consistent across all classifiers tested, highlighting the complexity of the interaction between noise-based defenses and different models. These results suggest that in some cases, noise-based defenses can inadvertently create an adversarial training loop beneficial to the RL attacker. Our study emphasizes the need for a more nuanced approach to defensive strategies in adversarial machine learning, particularly in safety-critical applications. It challenges the assumption that randomness universally enhances defense against evasion attacks and highlights the importance of considering adaptive, RL-based attackers when designing robust defense mechanisms.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Placed in Adversarial robustness and RL in finance, one of 15 directions in Deep reinforcement learning methods. Builds on 11 of its 20 references that are in the map. First posted on arXiv 31 Oct 2024.

Builds on

11 of 20 references.

References · 11

  1. 2015Deep Residual Learning for Image Recognitioncited by 52,096
  2. 2017Proximal Policy Optimization Algorithmscited by 15,215
  3. 2014Explaining and Harnessing Adversarial Examplescited by 7,364
  4. 2016Densely Connected Convolutional Networkscited by 6,720
  5. 2017Towards Deep Learning Models Resistant to Adversarial Attackscited by 6,415
  6. 2018MobileNetV2: Inverted Residuals and Linear Bottleneckscited by 4,795
  7. 2016Adversarial Machine Learning at Scalecited by 936
  8. 2017One pixel attack for fooling deep neural networkscited by 582
  9. 2018Certified Robustness to Adversarial Examples with Differential Privacycited by 468
  10. 2017Towards Robust Neural Networks via Random Self-ensemblecited by 159
  11. 2020PatchAttack: A Black-box Texture-based Attack with Reinforcement Learningcited by 67

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2025Adversarial Agents: Black-Box Evasion Attacks with Reinforcement Learning
  2. 2022Friendly Noise against Adversarial Noise: A Powerful Defense against Data Poisoning Attacks
  3. 2024Towards Robust Policy: Enhancing Offline Reinforcement Learning with Adversarial Attacks and Defenses
  4. 2026Reinforcement Learning Disrupts Gradient-Based Adversarial Optimization
  5. 2025Towards Robust Deep Reinforcement Learning against Environmental State Perturbation
  6. 2024Robust Deep Reinforcement Learning Through Adversarial Attacks and Training : A Survey
  7. 2024Let the Noise Speak: Harnessing Noise for a Unified Defense Against Adversarial and Backdoor Attacks
  8. 2023Defenses in Adversarial Machine Learning: A Survey