Concept Based Classification On Awa2
Metrics
Concept Accuracy (%)
Task Accuracy (%)
Results
Performance results of various models on this benchmark
| Paper Title | |||
|---|---|---|---|
| EQ-CBM (ResNet-34) | 99.129 | 95.965 | EQ-CBM: A Probabilistic Concept Bottleneck with Energy-based Models and Quantized Vectors |
| CGEM (ResNet-34) | 93.68 | 94.63 | Concept Graph Embedding Models for Enhanced Accuracy and Interpretability |
0 of 2 row(s) selected.