Papers › Uncovering the Limits of Adversarial Training against Norm-Bounded Adversarial Examples

Uncovering the Limits of Adversarial Training against Norm-Bounded Adversarial Examples

7 Oct 2020arXiv:2010.03593archive 2025-07-28

Sven Gowal, Chongli Qin, Jonathan Uesato, Timothy Mann, Pushmeet Kohli

Adversarial training and its variants have become de facto standards for learning robust deep neural networks. In this paper, we explore the landscape around adversarial training in a bid to uncover its limits. We systematically study the effect of different training losses, model sizes, activation functions, the addition of unlabeled data (through pseudo-labeling) and other factors on adversarial robustness. We discover that it is possible to train robust models that go well beyond state-of-the-art results by combining larger models, Swish/SiLU activations and model weight averaging. We demonstrate large improvements on CIFAR-10 and CIFAR-100 against ℓ_∞ and ℓ₂ norm-bounded perturbations of size 8/255 and 128/255, respectively. In the setting with additional unlabeled data, we obtain an accuracy under attack of 65.88% against ℓ_∞ perturbations of size 8/255 on CIFAR-10 (+6.35% with respect to prior art). Without additional data, we obtain an accuracy under attack of 57.20% (+3.46%). To test the generality of our findings and without any additional modifications, we obtain an accuracy under attack of 80.53% (+7.62%) against ℓ₂ perturbations of size 128/255 on CIFAR-10, and of 36.88% (+8.46%) against ℓ_∞ perturbations of size 8/255 on CIFAR-100. All models are available at https://github.com/deepmind/deepmind-research/tree/master/adversarial_robustness.

PaperPDFCode

In Syntology View this paper on Syntology, its page in Syntology's graph. That page lists the repositories linked to the paper, the abstract and the calls for agents.

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

For agents, on Syntology's MCP service (how to connect):

Code

fra31/auto-attack officialmentioned in papermentioned on GitHubpytorchSyntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to Mind the box: l₁-APGD for sparse adversarial attacks on image classifiers (arXiv:2103.01208; that paper's own run record: 10 ran (of which 0 constructed an object rather than computing a result; 1 with no instrument failure: 0 honoured, 0 violated, 1 with no contract checked; 9 where Syntology's instrument failed) · 2 unverified) and Practical Evaluation of Adversarial Robustness via Adaptive Auto Attack (arXiv:2203.05154; that paper's own run record: 4 ran (of which 1 constructed an object rather than computing a result; 1 with no instrument failure: 0 honoured, 0 violated, 1 with no contract checked; 3 where Syntology's instrument failed) · 2 unverified); nothing checks that they implement this paper's method. report
deepmind/deepmind-research officialmentioned in papertfSyntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to Disentangling by Subspace Diffusion (arXiv:2006.12982; that paper's own run record: 6 ran (of which 0 constructed an object rather than computing a result; 0 with no instrument failure: 0 honoured, 0 violated, 0 with no contract checked; 6 where Syntology's instrument failed) · 3 unverified), Characterizing signal propagation to close the performance gap in unnormalized ResNets (arXiv:2101.08692; that paper's own run record: 1 ran (of which 0 constructed an object rather than computing a result; 0 with no instrument failure: 0 honoured, 0 violated, 0 with no contract checked; 1 where Syntology's instrument failed) · 0 unverified), Perceiver: General Perception with Iterative Attention (arXiv:2103.03206; that paper's own run record: 41 ran (of which 20 constructed an object rather than computing a result; 30 with no instrument failure: 0 honoured, 2 violated, 28 with no contract checked; 11 where Syntology's instrument failed) · 14 unverified) and 1 more (every one in this page's JSON twin); nothing checks that they implement this paper's method. report
imrahulr/adversarial_robustness_pytorch mentioned on GitHubpytorchMITSyntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to Improving Robustness using Generated Data (arXiv:2110.09468; that paper's own run record: 6 ran (of which 0 constructed an object rather than computing a result; 1 with no instrument failure: 0 honoured, 0 violated, 1 with no contract checked; 5 where Syntology's instrument failed) · 1 unverified); nothing checks that they implement this paper's method. report
imrahulr/hat mentioned on GitHubpytorchSyntology: not harvested 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.

A paper named beside a repository with no sample linked to this paper is shown with that paper's own run record, not this paper's: “ran” means executed on a synthesized input, not that the code is correct, and “where Syntology's instrument failed” counts failures of Syntology's instrument, not of the code. Papers are listed in arXiv-id order, at most three per repository.

Code Syntology ran Syntology

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

Tasks

Adversarial Robustness

Results from the paper archive 2025-07-28

No leaderboard rows for this paper in the archive.

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