Research map / Adversarial robustness and security / Certified and empirical robustness
More Data Can Expand The Generalization Gap Between Adversarially Robust and Standard Models
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 |
|---|---|---|---|
| Increase training sample size in the strong-adversary regime | Gaussian classification model · Gaussian classification model | cross generalization gap | Hurt |
| Increase training sample size in the weak-adversary regime, past the threshold | Gaussian classification model · Gaussian classification model | cross generalization gap | Helped |
| Increase training sample size before the critical point | Gaussian classification model · Gaussian classification model | cross generalization gap | Hurt |
| Increase training sample size in the weak-adversary regime, past the threshold | Bernoulli classification model · Bernoulli classification model | cross generalization gap | Helped |
| Increase training sample size in the strong-adversary regime | Bernoulli classification model · Bernoulli classification model | cross generalization gap | Hurt |
| Increase training sample size | Linear regression with standard-normal inputs · Linear regression, x ~ N(0,1) | scaled cross generalization gap | Hurt |
| Increase training sample size under a weak adversary | Linear regression with shifted-Poisson inputs · Linear regression, x ~ Pois(5) + 1, small ε | scaled cross generalization gap | Helped |
| Increase training sample size under a strong adversary | Linear regression with shifted-Poisson inputs · Linear regression, x ~ Pois(5) + 1, large ε | scaled cross generalization gap | Hurt |
Closest papers by meaning
- Adversarially Robust Generalization Requires More Data (NeurIPS 2018 · similarity 0.91)
- Improving Adversarial Robustness by Putting More Regularizations on Less Robust Samples (ICML 2023 · similarity 0.91)
- Attacks Which Do Not Kill Training Make Adversarial Learning Stronger (ICML 2020 · similarity 0.90)
- Why adversarial training can hurt robust accuracy (ICLR 2023 · similarity 0.90)
- Adversarial Training Can Provably Improve Robustness: Theoretical Analysis of Feature Learning Process Under Structured Data (ICLR 2025 · similarity 0.89)
- A Closer Look at Curriculum Adversarial Training: From an Online Perspective (AAAI 2024 · 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.