Research map / Adversarial robustness and security / Certified and empirical robustness
On the Tightness of Semidefinite Relaxations for Certifying Robustness to Adversarial Examples
Richard Zhang
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 changed | Where | Measured | Effect |
|---|---|---|---|
| Increase network depth from one layer to two or more layers | CNN with the SDP relaxation solved by BM2 · CNN verification instances | numerical rank-1 success rate: 100% -> 0% | Hurt |
| Use an intermediate radius ρ in the stated loose interval | SDP relaxation for the original adversarial attack problem on dense-1 · dense-1 | SDP tightness: 0.4≤ρ≤60 | Hurt |
| Use a radius in either reported tight interval | SDP relaxation for the original adversarial attack problem on dense-1 · dense-1 | SDP tightness: 0.07≤ρ≤0.4 and ρ≥60 | Improved |
| Use a radius below the reported feasibility threshold | SDP relaxation for the original adversarial attack problem on dense-1 · dense-1 | SDP feasibility: ρ<0.07 | Hurt |
Closest papers by meaning
- Provably Cost-Sensitive Adversarial Defense via Randomized Smoothing (ICML 2025 · similarity 0.89)
- Expressive Losses for Verified Robustness via Convex Combinations (ICLR 2024 · similarity 0.88)
- Learning Better Certified Models from Empirically-Robust Teachers (NeurIPS 2026 · similarity 0.88)
- Probably Approximately Global Robustness Certification (ICML 2025 · similarity 0.88)
- Et Tu Certifications: Robustness Certificates Yield Better Adversarial Examples (ICML 2024 · similarity 0.88)
- GaLileo: General Linear Relaxation Framework for Tightening Robustness Certification of Transformers (AAAI 2024 · similarity 0.88)
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.
