Research map / Adversarial robustness and security / Certified and empirical robustness
Semidefinite relaxations for certifying robustness to adversarial examples
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 changed | Where | Measured | Effect |
|---|---|---|---|
| Use SDP-cert instead of Grad-cert | Grad-NN · MNIST, ℓ∞ ε=0.1 | fraction of non-certified examples: 35% -> 20% | Helped |
| Use SDP-cert instead of LP-cert | Grad-NN · MNIST, ℓ∞ ε=0.1 | fraction of non-certified examples: 97% -> 20% | Helped |
| Use SDP-cert instead of LP-cert | LP-NN · MNIST, ℓ∞ ε=0.1 | fraction of non-certified examples: 22% -> 20% | Helped |
| Use SDP-cert instead of Grad-cert | LP-NN · MNIST, ℓ∞ ε=0.1 | fraction of non-certified examples: 93% -> 20% | Helped |
| Use SDP-cert instead of LP-cert | PGD-NN · MNIST, ℓ∞ ε=0.1 | fraction of non-certified examples: 100% -> 18% | Helped |
| Add constraints relating linear and quadratic terms at intermediate layers (ablation) | SDP-cert · MNIST networks, ℓ∞ ε=0.1 | fraction of non-certified examples | Helped |
Closest papers by meaning
- Expressive Losses for Verified Robustness via Convex Combinations (ICLR 2024 · similarity 0.90)
- Learning Better Certified Models from Empirically-Robust Teachers (NeurIPS 2026 · similarity 0.90)
- On the Scalability of Certified Adversarial Robustness with Generated Data (NeurIPS 2024 · similarity 0.89)
- Provably Cost-Sensitive Adversarial Defense via Randomized Smoothing (ICML 2025 · similarity 0.89)
- Probably Approximately Global Robustness Certification (ICML 2025 · similarity 0.89)
- Adversarial Training Should Be Cast as a Non-Zero-Sum Game (ICLR 2024 · similarity 0.89)
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.