Papers › Evaluating the Robustness of Neural Networks: An Extreme Value Theory Approach

Evaluating the Robustness of Neural Networks: An Extreme Value Theory Approach

31 Jan 2018ICLR 2018 1arXiv:1801.10578archive 2025-07-28

Tsui-Wei Weng, huan zhang, Pin-Yu Chen, Jin-Feng Yi, Dong Su, Yupeng Gao, Cho-Jui Hsieh, Luca Daniel

The robustness of neural networks to adversarial examples has received great attention due to security implications. Despite various attack approaches to crafting visually imperceptible adversarial examples, little has been developed towards a comprehensive measure of robustness. In this paper, we provide a theoretical justification for converting robustness analysis into a local Lipschitz constant estimation problem, and propose to use the Extreme Value Theory for efficient evaluation. Our analysis yields a novel robustness metric called CLEVER, which is short for Cross Lipschitz Extreme Value for nEtwork Robustness. The proposed CLEVER score is attack-agnostic and computationally feasible for large neural networks. Experimental results on various networks, including ResNet, Inception-v3 and MobileNet, show that (i) CLEVER is aligned with the robustness indication measured by the ℓ₂ and ℓ_∞ norms of adversarial examples from powerful attacks, and (ii) defended networks using defensive distillation or bounded ReLU indeed achieve better CLEVER scores. To the best of our knowledge, CLEVER is the first attack-independent robustness metric that can be applied to any neural network classifier.

PaperPDFConference PDFCode

In Syntology Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

Code

huanzhang12/CLEVER officialmentioned in papertf report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

Not run by Syntology. Nothing on this page verifies that the listed code works.

Results from the paper archive 2025-07-28

No leaderboard rows for this paper in the archive.

Methods

1x1 ConvolutionAuxiliary ClassifierAverage PoolingBatch NormalizationBottleneck Residual BlockConvolutionDense ConnectionsDropoutGlobal Average PoolingInception-v3Inception-v3 ModuleKaiming InitializationLabel SmoothingMax PoolingRMSPropReLUResidual BlockResidual ConnectionSoftmax

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections