Towards Evaluating the Robustness of Neural Networks
Reviewed by Pithpith:7VL3EJT6open to challenge →
read the original abstract
Neural networks provide state-of-the-art results for most machine learning tasks. Unfortunately, neural networks are vulnerable to adversarial examples: given an input $x$ and any target classification $t$, it is possible to find a new input $x'$ that is similar to $x$ but classified as $t$. This makes it difficult to apply neural networks in security-critical areas. Defensive distillation is a recently proposed approach that can take an arbitrary neural network, and increase its robustness, reducing the success rate of current attacks' ability to find adversarial examples from $95\%$ to $0.5\%$. In this paper, we demonstrate that defensive distillation does not significantly increase the robustness of neural networks by introducing three new attack algorithms that are successful on both distilled and undistilled neural networks with $100\%$ probability. Our attacks are tailored to three distance metrics used previously in the literature, and when compared to previous adversarial example generation algorithms, our attacks are often much more effective (and never worse). Furthermore, we propose using high-confidence adversarial examples in a simple transferability test we show can also be used to break defensive distillation. We hope our attacks will be used as a benchmark in future defense attempts to create neural networks that resist adversarial examples.
This paper has not been read by Pith yet.
Forward citations
Cited by 10 Pith papers
-
BadSKP: Backdoor Attacks on Knowledge Graph-Enhanced LLMs with Soft Prompts
BadSKP poisons graph node embeddings to steer soft prompts in KG-enhanced LLMs, achieving high attack success rates where text-channel backdoors fail due to semantic anchoring.
-
Stateful Detection of Black-Box Adversarial Attacks
The paper argues for stateful defenses over stateless ones to detect adversarial example generation via query history and introduces query blinding as a counter-attack.
-
Uncovering Hidden Systematics in Neural Network Models for High Energy Physics
Neural networks for HEP tasks can be fooled at significant rates by subtle perturbations inside uncertainty envelopes, revealing hidden systematics not captured by conventional methods.
-
Transferable Physical-World Adversarial Patches Against Pedestrian Detection Models
TriPatch generates transferable physical adversarial patches via multi-stage triplet loss, appearance consistency, and data augmentation to achieve higher attack success rates on pedestrian detectors than prior methods.
-
SyncBreaker:Stage-Aware Multimodal Adversarial Attacks on Audio-Driven Talking Head Generation
SyncBreaker jointly attacks image and audio streams with Multi-Interval Sampling and Cross-Attention Fooling to degrade speech-driven talking head generation more than single-modality baselines.
-
Memory Efficient Full-gradient Attacks (MEFA) Framework for Adversarial Defense Evaluations
MEFA enables exact full-gradient white-box attacks on iterative stochastic purification defenses like diffusion and Langevin EBMs by trading recomputation for lower memory, revealing vulnerabilities missed by approxim...
-
Auto-ART: Structured Literature Synthesis and Automated Adversarial Robustness Testing
Auto-ART delivers the first structured synthesis of adversarial robustness consensus plus an executable multi-norm testing framework that flags gradient masking in 92% of cases on RobustBench and reveals a 23.5 pp rob...
-
Why Blocking Targeted Adversarial Perturbations Impairs the Ability to Learn
Defensive distillation blocks non-targeted adversarial attacks but cannot block targeted ones without preventing the network from learning via its input gradient.
-
Affine Disentangled GAN for Interpretable and Robust AV Perception
ADIS-GAN disentangles affine transformations in a GAN to achieve over 98% classification accuracy on MNIST within 30 degrees rotation and over 90% under FGSM and PGD attacks while generating rotation and scaling factors.
-
Measuring the Transferability of Adversarial Examples
Empirical measurement of adversarial example transferability between VGG and Inception model classes with methodological refinements to attack strength selection, perturbation clipping, and evaluation via SSIM.
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.