Towards Deep Learning Models Resistant to Adversarial Attacks

Madry, A., Makelov, A., Schmidt, L., Tsipras, D., & Vladu, A. (2018). Towards Deep Learning Models Resistant to Adversarial Attacks. International Conference on Learning Representations.

Type

Preprint

Intellectual base

Grey literature

Year

2018

Link

https://arxiv.org/abs/1706.06083

Cited from

security_intelligence

Cited by

2 citations across 2 artifacts.

Each citation links to the sentence it supports in the citing article.

Primes

• Adversarial Boundary Navigation
  • The adversarial-examples apparatus — a minimal perturbation within a budget that preserves semantics while crossing a classifier boundary — is the same apparatus as minimal corporate restructuring within a tax-code budget that preserves economic substance while crossing a safe-harbour boundary, and the defensive techniques map across too: adversarial training, ensemble defence, and certified robustness have direct analogues in anti-abuse doctrines, economic-substance tests, and multi-prong rules.

    This sourceEstablishes adversarial training as a robust-optimization defense; supports the defensive-technique mapping (adversarial training, ensemble, certified robustness).

Mechanisms

• Weakened Adversarial Example Set
  • In machine learning this is the basis of adversarial training.

    This sourceTrains machine-learning models on adversarially perturbed examples to improve robustness against such attacks.

Verification

Does it exist? Not checked yet. This entry carries no identifier to resolve. It was extracted from the citation as written in the article, normalized, and deduplicated against the rest of the registry.

Does it back the claim? Not recorded. Neither this nor any other of the 2 citations of this work carries a recorded support check.

Support is checked per citation rather than per work — the same source can be cited soundly in one article and wrongly in another. Per-citation recording began recently, so a citation with no recorded check is a gap in the record rather than evidence it went unchecked.

See how references were verified.

Registry ID ref:f3dab76ceb4e · see in the full table