Research map / Adversarial robustness and security / Certified and empirical robustness
Demystifying Causal Features on Adversarial Examples and Causal Inoculation for Robust Network by Adversarial Instrumental Variable Regression
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 |
|---|---|---|---|
| Add CAFE causal inoculation to HELP | HELP with ResNet-18 · CIFAR-10, AA | accuracy (%): 47.4 -> 48.5 | Helped |
| Add CAFE causal inoculation to HELP | HELP with ResNet-18 · SVHN, AA | accuracy (%): 52.0 -> 52.8 | Helped |
| Add CAFE causal inoculation to HELP | HELP with ResNet-18 · Tiny-ImageNet, AA | accuracy (%): 25.7 -> 27.3 | Helped |
| Add CAFE causal inoculation to ADV | ADV with VGG-16 · SVHN, FGSM | accuracy (%): 64.8 -> 67.0 | Helped |
| Add CAFE causal inoculation to defense baselines | ADV, TRADES, MART, AWP, and HELP · CIFAR-10, SVHN, and Tiny-ImageNet under adversarial attacks | accuracy (%) | Helped |
| Add CAFE causal inoculation to HELP | HELP with ResNet-18 · CIFAR-10, natural | accuracy (%): 80.5 -> 80.6 | No clear effect |
| Add CAFE causal inoculation to TRADES | TRADES with VGG-16 · CIFAR-10, natural | accuracy (%): 79.5 -> 77.0 | Hurt |
| Add the localized Rademacher regularizer to the test function (ablation) | AMR-GMM causal inference · Two datasets; VGG-16 | Rademacher distance | Helped |
| Add the localized Rademacher regularizer to the test function (ablation) | AMR-GMM causal inference · Two datasets; VGG-16 | prediction imbalance ratio | Helped |
Closest papers by meaning
- Adversarial Robustness Through the Lens of Causality (ICLR 2022 · similarity 0.92)
- Where and How to Attack? A Causality-Inspired Recipe for Generating Counterfactual Adversarial Examples (AAAI 2024 · similarity 0.90)
- Certified Causal Defense with Generalizable Robustness (AAAI 2025 · similarity 0.88)
- Adversarial Training Can Provably Improve Robustness: Theoretical Analysis of Feature Learning Process Under Structured Data (ICLR 2025 · similarity 0.87)
- Precise Accuracy / Robustness Tradeoffs in Regression: Case of General Norms (ICML 2024 · similarity 0.87)
- Causal Adversarial Perturbations for Individual Fairness and Robustness in Heterogeneous Data Spaces (AAAI 2024 · similarity 0.87)
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.