Research map / Adversarial robustness and security / Certified and empirical robustness
Towards Optimal Randomized Strategies in Adversarial Example Game
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 FRAT with 4 models instead of SAT | CIFAR-10 classifier · CIFAR-10 test | APGD-DLR accuracy: 48.1 -> 50.1 | Helped |
| Use FRAT with 4 models instead of SAT | CIFAR-10 classifier · CIFAR-10 test | natural accuracy: 80.6 -> 81.6 | Helped |
| Use FRAT with 4 models instead of SAT | CIFAR-10 classifier · CIFAR-10 test | APGD-CE accuracy: 49.7 -> 52.1 | Helped |
| Use FRAT with 4 models instead of SAT | CIFAR-10 classifier · CIFAR-10 test | APGD-CE & APGD-DLR accuracy: 47.6 -> 49.4 | Helped |
| Use FRAT with 4 models instead of SAT | CIFAR-100 classifier · CIFAR-100 test | natural accuracy: 54.1 -> 55.6 | Helped |
| Use FRAT with 4 models instead of SAT | CIFAR-100 classifier · CIFAR-100 test | APGD-DLR accuracy: 23.4 -> 26.1 | Helped |
| Use FRAT with 4 models instead of SAT | CIFAR-100 classifier · CIFAR-100 test | APGD-CE & APGD-DLR accuracy: 23.2 -> 25.5 | Helped |
| Use FRAT with 4 models instead of SAT | CIFAR-100 classifier · CIFAR-100 test | APGD-CE accuracy: 26.5 -> 28.7 | Helped |
| Use BAT instead of SAT | CIFAR-10 classifier · CIFAR-10 test | APGD-CE accuracy: 49.7 -> 53.3 | Helped |
| Use BAT instead of SAT | CIFAR-10 classifier · CIFAR-10 test | APGD-DLR accuracy: 48.1 -> 45.2 | Hurt |
Closest papers by meaning
- Randomization matters How to defend against strong adversarial attacks (ICML 2020 · similarity 0.89)
- Mixed Nash Equilibria in the Adversarial Examples Game (ICML 2021 · similarity 0.89)
- Adversarial Example Games (NeurIPS 2020 · similarity 0.89)
- No Free Lunch Theorem and Black-Box Complexity Analysis for Adversarial Optimisation (NeurIPS 2024 · similarity 0.85)
- Adversarial Robustness via Deformable Convolution with Stochasticity (ICML 2025 · similarity 0.84)
- Learning Randomized Algorithms with Transformers (ICLR 2025 · similarity 0.84)
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.