Research map / Label-efficient and robust learning / Recognition and long-tail
Label-free Concept Bottleneck Models
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 |
|---|---|---|---|
| Replace a standard network with a label-free concept bottleneck model | ResNet-50 · ImageNet | accuracy: 76.13% -> 71.95% | Hurt |
| Use Label-free CBM instead of P-CBM | CLIP(RN50) · CIFAR-10 | accuracy: 70.50%* -> 86.40% | Helped |
| Use Label-free CBM instead of P-CBM | CLIP(RN50) · CIFAR-100 | accuracy: 43.20%* -> 65.13% | Helped |
| Use Label-free CBM instead of P-CBM | ResNet-18 · CUB200 | accuracy: 59.60%* -> 74.31% | Helped |
| Use Label-free CBM instead of P-CBM (CLIP) | CLIP(RN50) · CIFAR-10 | accuracy: 84.50%* -> 86.40% | Helped |
| Use Label-free CBM instead of P-CBM (CLIP) | CLIP(RN50) · CIFAR-100 | accuracy: 56.00%* -> 65.13% | Helped |
| Use Label-free CBM instead of a standard sparse model | CLIP(RN50) · CIFAR-10 | accuracy: 82.96% -> 86.40% | Helped |
| Use Label-free CBM instead of a standard sparse model | CLIP(RN50) · CIFAR-100 | accuracy: 58.34% -> 65.13% | Helped |
| Use Label-free CBM instead of a standard sparse model | ResNet-18 · CUB200 | accuracy: 75.96% -> 74.31% | Hurt |
| Use Label-free CBM instead of a standard sparse model | ResNet-50 · Places365 | accuracy: 38.46% -> 43.68% | Helped |
| Use Label-free CBM instead of a standard sparse model | ResNet-50 · ImageNet | accuracy: 74.35% -> 71.95% | Hurt |
| Manually edit final-layer weights to correct identified errors | Label-free CBM · ImageNet validation | accuracy: 71.98% to 72.02% | Helped |
Closest papers by meaning
- Hybrid Concept Bottleneck Models (CVPR 2025 · similarity 0.90)
- Energy-Based Concept Bottleneck Models: Unifying Prediction, Concept Intervention, and Probabilistic Interpretations (ICLR 2024 · similarity 0.90)
- VLG-CBM: Training Concept Bottleneck Models with Vision-Language Guidance (NeurIPS 2024 · similarity 0.90)
- Concepts Worth Having: Refining VLM-Guided Concept Bottleneck Models with Minimal Annotations (NeurIPS 2026 · similarity 0.89)
- Interpretable Generative Models through Post-hoc Concept Bottlenecks (CVPR 2025 · similarity 0.89)
- DCBM: Data-Efficient Visual Concept Bottleneck Models (ICML 2025 · similarity 0.89)
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.