AI Paper Atlas

Research map / Learning theory and algorithms / Combinatorial and search optimization

A Semidefinite Relaxation Based Branch-and-Bound Method for Tight Neural Network Verification

Jianglin Lan, Benedikt Brückner, Alessio Lomuscio

AAAI 2023 · Learning theory and algorithms › Combinatorial and search optimization · approach: Semidefinite relaxations for verification · 2 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
Add eigenvector-based linear constraints and branch-and-bound to LayerSDPLayerSDP · MNIST MLP-Advverified robustness: 91 -> 92Improved
Add eigenvector-based linear constraints and branch-and-bound to LayerSDPLayerSDP · MNIST MLP-LPverified robustness: 80 -> 80No clear effect
Add eigenvector-based linear constraints and branch-and-bound to LayerSDPLayerSDP · MNIST MLP-SDPverified robustness: 84 -> 84No clear effect
Add eigenvector-based linear constraints and branch-and-bound to LayerSDPLayerSDP · MNIST MLP-6×100verified robustness: 75 -> 80Improved
Add eigenvector-based linear constraints and branch-and-bound to LayerSDPLayerSDP · MNIST MLP-9×100verified robustness: 35 -> 47Improved
Use BaBSDP instead of β-CROWNBaBSDP · MNIST MLP-9×100verified robustness: 69 -> 47Hurt
Add eigenvector-based linear constraints and branch-and-bound to LayerSDPLayerSDP · MNIST MLP-6×200verified robustness: 92 -> 93Improved
Add eigenvector-based linear constraints and branch-and-bound to LayerSDPLayerSDP · CIFAR10 CF-3×20verified robustness: 17 -> 18Improved

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