Research map / Learning theory and algorithms / Combinatorial and search optimization
Error Slice Discovery via Manifold Compactness
Han Yu, Hao Zou, Jiashuo Liu, Renzhe Xu, Yue He, Xingxuan Zhang, Peng Cui
Read the paperCheck your idea against 137,217 papers
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 |
|---|---|---|---|
| replace Domino/Spotlight/PlaneSpot with MCSD (joint risk and manifold compactness QP) | error slice discovery · dcbench (correlation, rare, noisy) | precision@10 (%): 36.2 -> 47.4 (correlation, vs Domino) | Improved |
| use MCSD instead of Domino on CelebA blond-hair category | error slice discovery · CelebA blond-hair category | manifold compactness: 6.07 -> 8.09 | Improved |
| MCSD vs Spotlight on slice accuracy in CelebA blond-hair | error slice discovery · CelebA blond-hair category | accuracy (%): 26.3 (Spotlight) vs 33.8 (MCSD) | Hurt |
| apply MCSD on BDD100K object detection (loss-based, not classification only) | error slice discovery · BDD100K pedestrian | manifold compactness: 2.05 -> 6.60 | Improved |
| MCSD vs Spotlight on BDD100K traffic-light slice AP | error slice discovery · BDD100K traffic light | AP (%): 46.3 (Spotlight) vs 57.3 (MCSD) | Hurt |
| use MCSD on CivilComments toxic category | error slice discovery · CivilComments toxic | accuracy (%): 48.6 -> 25.2 (vs Spotlight; lower is better) | Improved |
| measure coherence by manifold compactness rather than variance across slice granularity | coherence metric · CelebA | coherence monotonicity | Improved |
Closest papers by meaning
- CB-SLICE: Concept-Based Interpretable Error Slice Discovery (ICML 2026 · similarity 0.86)
- Domino: Discovering Systematic Errors with Cross-Modal Embeddings (ICLR 2022 · similarity 0.86)
- Recovering Manifold Structure Using Ollivier Ricci Curvature (ICLR 2025 · similarity 0.83)
- HiBug2: Efficient and Interpretable Error Slice Discovery for Comprehensive Model Debugging (ICLR 2025 · similarity 0.83)
- Avoiding spurious sharpness minimization broadens applicability of SAM (ICML 2025 · similarity 0.82)
- Hardness of Learning Neural Networks under the Manifold Hypothesis (NeurIPS 2024 · similarity 0.82)
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.
