Average across datasets
Benchmarks
| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Faithful Calibration | Average across datasets (PQA, SA, SQA, HE, MMLU, SQ, MT, UM, AC, SG) | cMFG*0.85 | 27 | |
| Image Classification | Average across datasets | Base Score89.05 | 15 | |
| Disease diagnosis | Average across datasets | AUC89.11 | 15 | |
| Conformal Image Classification | Average across 11 datasets CLIP ViT-B/16 features (test) | Accuracy85.6 | 9 | |
| Image Classification | Average across 11 datasets (Aircraft, CIFAR10, CIFAR100, CUB200, DTD, Flower102, Food101, HAM10000, ImageNet, Resisc45, UCF101) | Avg ACC87.4 | 9 | |
| Time Series Denoising | Average Across Datasets Scenario 3 | RMSE0.0568 | 8 | |
| Time Series Denoising | Average Across Datasets Scenario 2 | RMSE0.0801 | 8 | |
| Time Series Denoising | Average Across Datasets Scenario 1 | RMSE0.0456 | 8 | |
| Average performance across 10 task types | Average across 13 datasets (test) | Avg. Accuracy75.8 | 8 | |
| Text-image retrieval | Average across datasets | Speedup3.5 | 5 | |
| Depth Completion | Average across datasets Hole | AbsREL1 | 4 |
Showing 11 of 11 rows