Research map / Label-efficient and robust learning / OOD and anomaly detection
Breaking Correlation Shift via Conditional Invariant Regularizer
Mingyang Yi, Ruoyu Wang, Jiacheng Sun, Zhenguo Li, Zhi-Ming Ma
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 |
|---|---|---|---|
| Regularize with CSV using observable spurious attributes (RCSV) | ResNet50 · CelebA | Avg accuracy (%): 91.6 -> 92.6 | Improved |
| Regularize with CSV using observable spurious attributes (RCSV) | ResNet50 · Waterbirds | Avg accuracy (%): 90.2 -> 90.5 | Improved |
| Regularize with CSV using observable spurious attributes (RCSV) | BERT · MultiNLI | Avg accuracy (%): 82.0 -> 82.6 | Improved |
| Regularize with CSV using observable spurious attributes (RCSV) | BERT · CivilComments | Avg accuracy (%): 82.2 -> 83.7 | Improved |
| Estimate and regularize CSV without observable spurious attributes (RCSV_U) | ResNet50 · CelebA | Avg accuracy (%): 85.5 -> 89.8 | Improved |
| Estimate and regularize CSV without observable spurious attributes (RCSV_U) | ResNet50 · Waterbirds | Avg accuracy (%): 84.3 -> 89.1 | Improved |
| Estimate and regularize CSV without observable spurious attributes (RCSV_U) | BERT · MultiNLI | Avg accuracy (%): 79.5 -> 80.0 | Improved |
| Estimate and regularize CSV without observable spurious attributes (RCSV_U) | BERT · CivilComments | Avg accuracy (%): 83.5 -> 81.7 | Hurt |
| Use observable spurious attributes for CSV estimation instead of the unobservable estimator | RCSV/RCSV_U · CelebA, Waterbirds, MultiNLI, CivilComments | Avg accuracy (%) | Improved |
| Use the stated inverse-frequency reweighted sampling strategy when estimating CSV (ablation) | RCSV and RCSV_U · OOD benchmarks | OOD generalization | Improved |
Closest papers by meaning
- Towards a Theoretical Framework of Out-of-Distribution Generalization (NeurIPS 2021 · similarity 0.90)
- Regularization Penalty Optimization for Addressing Data Quality Variance in OoD Algorithms (AAAI 2022 · similarity 0.88)
- Aggregation Hides Out-of-Distribution Generalization Failures from Spurious Correlations (NeurIPS 2025 · similarity 0.88)
- Certifiable Out-of-Distribution Generalization (AAAI 2023 · similarity 0.88)
- Towards Robust Out-of-Distribution Generalization Bounds via Sharpness (ICLR 2024 · similarity 0.87)
- The Best of Both Worlds: On the Dilemma of Out-of-distribution Detection (NeurIPS 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.
