mlsanityData to 7 Oct 2026

On the tightness of linear relaxation based robustness certification methods

arXiv 2210.00178 ↗First posted 1 Oct 2022cs.LGcited by 0

Not on the map. Why?

Abstract (arXiv)

There has been a rapid development and interest in adversarial training and defenses in the machine learning community in the recent years. One line of research focuses on improving the performance and efficiency of adversarial robustness certificates for neural networks citegowal:19, wongzico:18, raghunathan:18, WengTowardsFC:18, wong:scalable:18, singh:convexbarrier:19, Huangetal:19, single-neuron-relax:20, Zhang2020TowardsSA. While each providing a certification to lower (or upper) bound the true distortion under adversarial attacks via relaxation, less studied was the tightness of relaxation. In this paper, we analyze a family of linear outer approximation based certificate methods via a meta algorithm, IBP-Lin. The aforementioned works often lack quantitative analysis to answer questions such as how does the performance of the certificate method depend on the network configuration and the choice of approximation parameters. Under our framework, we make a first attempt at answering these questions, which reveals that the tightness of linear approximation based certification can depend heavily on the configuration of the trained networks.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Builds on 18 of its 21 references that are in the map. First posted on arXiv 1 Oct 2022.

Builds on

18 of 21 references.

References · 18

  1. 2014Explaining and Harnessing Adversarial Examplescited by 7,287
  2. 2013Intriguing properties of neural networkscited by 5,114
  3. 2016Towards Evaluating the Robustness of Neural Networkscited by 2,976
  4. 2021Convex optimizationcited by 2,479
  5. 2017Reluplex: An Efficient SMT Solver for Verifying Deep Neural Networkscited by 739
  6. 2017Provable defenses against adversarial examples via the convex outer adversarial polytopecited by 564
  7. 2018Efficient Neural Network Robustness Certification with General Activation Functionscited by 439
  8. 2018Towards Fast Computation of Certified Robustness for ReLU Networkscited by 254
  9. 2019Towards Stable and Efficient Training of Verifiably Robust Neural Networkscited by 204
  10. 2018Semidefinite relaxations for certifying robustness to adversarial examplescited by 178
  11. 2018Scaling provable adversarial defensescited by 141
  12. 2019A Convex Relaxation Barrier to Tight Robustness Verification of Neural Networkscited by 124
6 more
  1. 2019Achieving Verified Robustness to Symbol Substitutions via Interval Bound Propagationcited by 87
  2. 2018CNN-Cert: An Efficient Framework for Certifying Robustness of Convolutional Neural Networkscited by 50
  3. 2020The Convex Relaxation Barrier, Revisited: Tightened Single-Neuron Relaxations for Neural Network Verificationcited by 50
  4. 2017Towards Proving the Adversarial Robustness of Deep Neural Networkscited by 26
  5. 2017Evaluating Robustness of Neural Networks with Mixed Integer Programmingcited by 14
  6. 2020On the Tightness of Semidefinite Relaxations for Certifying Robustness to Adversarial Examplescited by 14

Built on by

Not cited yet.