Clear
Benchmarks
| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Out-of-Distribution Detection | CLEAR10 ID | AUROC (COCO)99.66 | 40 | |
| Out-of-Distribution Detection | CLEAR100 ID | AUROC (COCO)97.37 | 40 | |
| Visual Question Answering | CLEAR 1.0 (Retain) | Accuracy70.9 | 32 | |
| Classification | CLEAR | Error Rate4 | 24 | |
| Multimodal Machine Unlearning | CLEAR | Forget VQA Accuracy32.4 | 16 | |
| Machine Unlearning | CLEAR (test 2) | Forget Accuracy44 | 16 | |
| Machine Unlearning | CLEAR (test 1) | Forget Accuracy42 | 16 | |
| Question Answering | CLEAR Real-world 1.0 | Acc94.7 | 16 | |
| Question Answering | CLEAR 1.0 (Retain) | R-L Score0.352 | 16 | |
| Question Answering | CLEAR Forget 1.0 | R-L Score0.367 | 16 | |
| Visual Question Answering | CLEAR Forget 1.0 | Accuracy34.2 | 16 | |
| Domain-incremental learning | CLEAR-10 | Accuracy67.67 | 13 | |
| Online Continual Self-Supervised Learning | CLEAR100 11 experiences (streaming online) | Final Accuracy51.5 | 9 | |
| Classification | CLEAR (Realworld) | Accuracy (Realworld)74.66 | 5 | |
| Text Generation | CLEAR 5% forget | Forget Gen.42 | 5 | |
| Classification | CLEAR 5% Forget | Accuracy (Forget Set)78.19 | 5 | |
| Bias Evaluation | CLEAR Bias | Age Performance82.9 | 5 | |
| Readability Assessment | CLEAR | Pearson Correlation0.735 | 4 | |
| Visual Question Answering | CLEAR Real QA | Accuracy (Aut)76.6 | 4 | |
| Identity Recognition | CLEAR (Retain) | Recall4.21 | 4 | |
| Identity Recognition | CLEAR (Forget) | Recall62 | 4 | |
| Depth Completion | Clear-Real (test) | RMSE0.041 | 4 | |
| Prediction | CLEAR Control Group | Time per 1000 Iterations287.85 | 3 | |
| Prediction | CLEAR (Treatment Group) | Time per 1000 Iterations232.64 | 3 | |
| Causal Reasoning | CLEAR | Accuracy60.5 | 3 |
Showing 25 of 30 rows