Research map / Adversarial robustness and security / Certified and empirical robustness
Improving Generalization of Adversarial Training via Robust Critical Fine-Tuning
Kaijie Zhu, Xixu Hu, Jindong Wang, Xing Xie, Ge Yang
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 |
|---|---|---|---|
| fine-tune the lowest-MRC (non-robust-critical) module and interpolate weights | Adversarial training · CIFAR10/CIFAR100/Tiny-ImageNet | Std accuracy: +1.21 / +2.02 / +1.87 avg | Improved |
| apply RiFT to adversarially trained model | Adversarial training · CIFAR10-C/CIFAR100-C/Tiny-ImageNet-C | OOD accuracy: +1.49 / +1.81 / +1.75 avg | Improved |
| apply RiFT to adversarially trained model | Adversarial training · CIFAR10/CIFAR100/Tiny-ImageNet | Adv accuracy (PGD-10): +0.02 / +0.02 / +0.08 avg | No clear effect |
| apply RiFT to TRADES, MART, AWP, SCORE | ResNet18 · CIFAR10/CIFAR100 | Std accuracy: SCORE CIFAR100 54.83 -> 57.63 | Improved |
| fine-tune non-robust-critical module instead of all layers (ablation) | RiFT · CIFAR10 (ResNet18) | Adv accuracy (PGD-10): 52.66 -> 53.65 | Improved |
| fine-tune robust-critical module instead of non-robust-critical module (ablation) | RiFT · CIFAR10 (ResNet18) | Adv accuracy (PGD-10): 53.65 -> 52.48 | Hurt |
| fine-tune last layer only (linear probing) (ablation) | RiFT · CIFAR10 (ResNet18) | Adv accuracy (PGD-10): 53.65 -> 52.75 | Hurt |
| fine-tune more non-robust-critical modules (top 2, 3, 5) instead of top 1 (ablation) | RiFT · CIFAR10 (ResNet18) | Adv accuracy (PGD-10): 53.65 -> 52.47 (top 2) | Hurt |
| apply RiFT to WRN34-10 on CIFAR10 | WRN34-10 · CIFAR10 | Std accuracy: 87.41 -> 87.89 | Improved |
Closest papers by meaning
- Adversarial Robustness: From Self-Supervised Pre-Training to Fine-Tuning (CVPR 2020 · similarity 0.91)
- Fantastic Robustness Measures: The Secrets of Robust Generalization (NeurIPS 2023 · similarity 0.91)
- Sparse Fine-Tuning for Parameter-Efficient Adversarial Training (NeurIPS 2026 · similarity 0.91)
- Boosting Adversarial Robustness with CLAT: Criticality Leveraged Adversarial Training (ICML 2025 · similarity 0.90)
- Adversarial Training Can Provably Improve Robustness: Theoretical Analysis of Feature Learning Process Under Structured Data (ICLR 2025 · similarity 0.89)
- Explaining the role of Intrinsic Dimensionality in Adversarial Training (ICML 2025 · 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.
