Research map / Adversarial robustness and security / Black-box and transfer attacks
Uncovering the Connections Between Adversarial Transferability and Knowledge Transferability
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 adversarial transferability metrics (alpha_1, alpha_2) of source models to predict knowledge transferability | CIFAR10-trained architectures transferred to STL10 · CIFAR10->STL10 model architectures | Pearson correlation between transfer loss and alpha_1 | Helped |
| Adversarial transferability metric (alpha_1) computed on source models for NLP transfer | BERT source models (MR, Yelp, AG, Fake) to IMDB · IMDB NLP | Pearson correlation with transfer confidence | Mixed |
| Vary source similarity to target distribution (0% to 100%) | Source models for CIFAR10->STL10 · Similarity-controlled source models | Pearson correlation between transfer loss and alpha_1 | No clear effect |
| Change attack method (PGD-L2, MI) and epsilon (2/255, 4/255) for adversarial transferability measurement (ablation) | Adversarial transferability metrics across architectures · CIFAR10->STL10 robustness ablation | Pearson correlation with knowledge transferability: Pearson between -0.40 and -0.49 across variants | No clear effect |
| Higher knowledge transferability of source model (more similar source data) corresponds to higher adversarial transferability | Source models from CIFAR10 subsets to STL10 vehicles target · Knowledge-to-adversarial transfer experiment | adversarial transferability (alpha_1, alpha_2) | Helped |
Closest papers by meaning
- A Unified Approach to Interpreting and Boosting Adversarial Transferability (ICLR 2021 · similarity 0.89)
- Transferability Bound Theory: Exploring Relationship between Adversarial Transferability and Flatness (NeurIPS 2024 · similarity 0.89)
- Enhancing Adversarial Transferability with Adversarial Weight Tuning (AAAI 2025 · similarity 0.88)
- Understanding Model Ensemble in Transferable Adversarial Attack (ICML 2025 · similarity 0.87)
- Boosting Adversarial Transferability through Augmentation in Hypothesis Space (CVPR 2025 · similarity 0.87)
- Enhancing Adversarial Transferability with Checkpoints of a Single Model's Training (CVPR 2025 · similarity 0.86)
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.