Research map / Probabilistic and scientific modeling / Molecules and proteins
Evidential Copula Concept Embedding Models
Yanjie Qiu, Xiaodong Yue, Xuhui Fan, Yufei Chen, Jie Shi, Wei Liu
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 |
|---|---|---|---|
| add Gaussian copula correlation to evidential CEM (ablation) | EC-CEM (evi-CEM) · AwA2 | Task accuracy: 88.87 -> 93.04 | Improved |
| add Gaussian copula correlation to evidential CEM (ablation) | EC-CEM (evi-CEM) · CUB | Task accuracy: 74.48 -> 76.19 | Improved |
| add Gaussian copula correlation to evidential CEM (ablation) | EC-CEM (evi-CEM) · SkinCon | Task accuracy: 77.69 -> 78.52 | Improved |
| add evidential regularization to CEM without copula (ablation) | evi-CEM · AwA2 | Task accuracy: 90.66 -> 88.87 | Hurt |
| add evidential regularization to CEM without copula (ablation) | evi-CEM · CUB | Task accuracy: 75.48 -> 74.48 | No clear effect |
| reduce available concepts (sparse concepts) (ablation) | EC-CEM vs AR-CBM · CUB | Task accuracy gap: 3.08% -> 57.24% | Improved |
| reduce available concepts (sparse concepts) (ablation) | EC-CEM vs AR-CBM · AwA2 | Task accuracy gap: 2.6% -> 31.96% | Improved |
| use evidential epistemic uncertainty for intervention selection instead of SCBM distance heuristic (ablation) | EC-CEM · CUB and AwA2 | Task F1 under intervention | Improved |
| add copula dependency modeling | EC-CEM · CUB | Concept ECE: 6.48 -> 4.39 | Improved |
| add copula dependency modeling | EC-CEM · AwA2 | CAS: 86.34 -> 91.38 | Improved |
| add copula dependency modeling | EC-CEM · SkinCon | CAS: 62.95 -> 63.17 | No clear effect |
| use learned copula correlation instead of by-chance permutation (ablation) | EC-CEM · CUB concept groups | Group Correction Score | Improved |
Closest papers by meaning
- Energy-Based Concept Bottleneck Models: Unifying Prediction, Concept Intervention, and Probabilistic Interpretations (ICLR 2024 · similarity 0.84)
- Sample-efficient Learning of Concepts with Theoretical Guarantees: from Data to Concepts without Interventions (NeurIPS 2025 · similarity 0.83)
- Hyperbolic Concept Bottleneck Models (NeurIPS 2026 · similarity 0.83)
- Learning Discrete Concepts in Latent Hierarchical Models (NeurIPS 2024 · similarity 0.83)
- Concepts Worth Having: Refining VLM-Guided Concept Bottleneck Models with Minimal Annotations (NeurIPS 2026 · similarity 0.83)
- Stochastic Concept Bottleneck Models (NeurIPS 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.
