AI Paper Atlas

Research map / Adversarial robustness and security / Certified and empirical robustness

Towards Reliable Learning for High Stakes Applications

Jinyang Gao, Junjie Yao, Yingxia Shao

AAAI 2019 · Adversarial robustness and security › Certified and empirical robustness · approach: Theory of adversarial risk · 13 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
use GALVE reliability R(x)*confidence instead of confidence alonereject-option ResNet-56 · CIFAR10error rate at 30% selectivity: 4.20 -> 1.90Improved
use GALVE instead of confidencereject-option ResNet-56 · SVHNerror rate at 30% selectivity: 1.05 -> 0.43Improved
use plain GAN discriminator instead of confidencereject-option ResNet-56 · CIFAR10error rate at 30% selectivity: 4.20 -> 3.4Improved
use entropy instead of confidencereject-option ResNet-56 · CIFAR10error rate at 30% selectivity: 4.20 -> 4.32Hurt
train R(x) on high-variance samples G(cz), c=1.1, instead of plain GAN discriminator (c=1.0) (ablation)GALVE · CIFAR10error rate at 30% selectivity: 3.4 -> 1.90Improved
scale factor c=1.2 instead of 1.1 (ablation)GALVE · CIFAR10error rate at 30% selectivity: 1.90 -> 2.67Hurt
scale factor c=0.9 (below 1) instead of c=1.0 (ablation)GALVE · CIFAR10error rate at 30% selectivity: 3.4 -> 3.27Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy