Research map / Label-efficient and robust learning / Test-time adaptation
Robustness to Spurious Correlations via Human Annotations
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 crowdsourced annotations as approximate unmeasured variables for UV-DRO | Logistic regression · MNIST rotated test set | accuracy: 43% -> 48% | Helped |
| Use oracle unmeasured variables for UV-DRO | Logistic regression · MNIST rotated test set | accuracy: 48% vs. 52% | Helped |
| Shuffle annotation distances (ablation) | UV-DRO with crowdsourced annotations · MNIST rotated test set | accuracy: 48% -> 42% | Hurt |
| Sample the unmeasured variable from c given x instead of c given x and y (ablation) | UV-DRO · Simulated medical diagnosis task | test loss | Hurt |
| Use crowdsourced annotations as approximate unmeasured variables for UV-DRO | Logistic regression · NYPD Police Stops, Brooklyn test set | accuracy: 60.5% -> 62% | Helped |
| Shuffle crowdsourced annotation information (ablation) | UV-DRO with crowdsourced annotations · NYPD Police Stops, Brooklyn test set | accuracy | No clear effect |
Closest papers by meaning
- Let Samples Speak: Mitigating Spurious Correlation by Exploiting the Clusterness of Samples (CVPR 2025 · similarity 0.84)
- Bayesian Inference for Correlated Human Experts and Classifiers (ICML 2025 · similarity 0.84)
- Learning from Noisy Labels via Conditional Distributionally Robust Optimization (NeurIPS 2024 · similarity 0.84)
- Mitigating Spurious Correlations via Disagreement Probability (NeurIPS 2024 · similarity 0.84)
- idSCD: Identifying Training Datasets through Semantic Correlation Descriptors (NeurIPS 2026 · similarity 0.83)
- Rethinking Robustness of Model Attributions (AAAI 2024 · similarity 0.83)
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.