Research map / Adversarial robustness and security / Certified and empirical robustness
Adversarial Example Games
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 AEG generator instead of best heuristic baseline (DI/MI/TID/SGM) for unseen-model attacks | AEG · CIFAR-10 known architecture NoBox | attack success rate (%): 84.0 -> 87.0 | Helped |
| use AEG instead of DI-Attack on MNIST known-architecture NoBox | AEG · MNIST known architecture NoBox | attack success rate (%): 89.5 -> 89.5 | No clear effect |
| use AEG instead of best baseline when transferring across architectures from VGG-16 source | AEG · CIFAR-10 cross-architecture transfer | error rate (%): 65.1 -> 97.5 (VGG-16 target) | Helped |
| use AEG instead of best baseline against ensemble adversarially trained CIFAR-10 models | AEG · CIFAR-10 robust targets | error rate (%): 33.1 -> 52.2 (RN-18_ens3) | Helped |
| use AEG instead of best baseline against Madry PGD-adversarially-trained models | AEG · MNIST Madry-Adv | error rate (%): 3.1 -> 5.9 | Helped |
| use AEG instead of gradient baselines (MI/DI/SGM) in the appendix average over architectures on CIFAR-10 | AEG · CIFAR-10 average over architectures (appendix) | test error rate (%): 99.9 -> 98.5 | Hurt |
| use AEG instead of baselines when source is Wide-ResNet (original AEG) | AEG · CIFAR-10 transfer from WR | error rate (%): 67.8 -> 40.8 (VGG-16 target) | Mixed |
| use AEG instead of whitebox/blackbox-query attacks transferred to unseen targets | AEG · MNIST NoBox average (appendix) | test error rate (%): 95.8 -> 98.9 | Helped |
Closest papers by meaning
- Are Adversarial Examples Created Equal? A Learnable Weighted Minimax Risk for Robustness under Non-uniform Attacks (AAAI 2021 · similarity 0.90)
- Are adversarial examples inevitable? (ICLR 2019 · similarity 0.89)
- On Adversarial Training without Perturbing all Examples (ICLR 2024 · similarity 0.89)
- One Man's Trash Is Another Man's Treasure: Resisting Adversarial Examples by Adversarial Examples (CVPR 2020 · similarity 0.89)
- Game-Theoretic Unlearnable Example Generator (AAAI 2024 · similarity 0.88)
- Boosting Adversarial Transferability through Augmentation in Hypothesis Space (CVPR 2025 · 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.