Research map / Adversarial robustness and security / Certified and empirical robustness
Stable Adversarial Learning under Distributional Shifts
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 SAL's learned anisotropic uncertainty set, weighting covariates by stability | Linear regression under selection bias · Selection bias simulations | Mean_Error: 0.475 -> 0.450; 0.464 -> 0.449; 0.477 -> 0.452; 0.460 -> 0.454; 0.464 -> 0.451; 0.459 -> 0.448 | Helped |
| Use SAL's learned anisotropic uncertainty set, weighting covariates by stability | Prediction under anti-causal shifts · Anti-causal effect, scenario 1, testing environments e5–e10 | prediction error: 0.382 -> 0.380; 0.420 -> 0.403; 0.444 -> 0.425; 0.461 -> 0.435; 0.478 -> 0.446; 0.504 -> 0.458 | Helped |
| Use SAL's learned anisotropic uncertainty set, weighting covariates by stability | Prediction under anti-causal shifts · Anti-causal effect, scenario 2, testing environments e4–e10 | prediction error: 0.328 -> 0.287; 0.343 -> 0.288; 0.358 -> 0.287; 0.365 -> 0.290; 0.374 -> 0.284; 0.377 -> 0.293; 0.394 -> 0.294 | Helped |
| Use SAL instead of IRM | Colored MNIST classifier · Colored MNIST test | Test Acc: 0.51 -> 0.57 | Helped |
| Use SAL instead of IRM | Real-world house-price regression · King County house sales, later periods | Mean_Error | Helped |
| Use SAL instead of IRM | Adult income classifier · Adult demographic environments | mis-classification rate | Helped |
Closest papers by meaning
- Certifying Some Distributional Robustness with Principled Adversarial Training (ICLR 2018 · similarity 0.88)
- Tolerant Algorithms for Learning with Arbitrary Covariate Shift (NeurIPS 2024 · similarity 0.87)
- Towards Robustness to Natural Variations and Distribution Shift (Student Abstract) (AAAI 2024 · similarity 0.86)
- Provable Robust Overfitting Mitigation in Wasserstein Distributionally Robust Optimization (ICLR 2025 · similarity 0.86)
- TAET: Two-Stage Adversarial Equalization Training on Long-Tailed Distributions (CVPR 2025 · similarity 0.86)
- Revisiting Adversarial Training Under Long-Tailed Distributions (CVPR 2024 · 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.