| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Open-set knowledge retrieval | KAMEL | Macro EM16.2 | 19 | |
| Model Editing | KAMEL multi-token targets (N=4) | Efficacy73 | 4 | |
| Model Editing | KAMEL multi-token targets (N=3) | Efficacy76 | 4 | |
| Model Editing | KAMEL multi-token targets N=2 | Efficacy88 | 4 | |
| Editing Efficacy | KAMEL 4 tokens | Editing Efficacy95 | 4 | |
| Editing Efficacy | KAMEL 3 tokens | Editing Efficacy96.5 | 4 | |
| Editing Efficacy | KAMEL 2 tokens | Editing Efficacy98.5 | 4 | |
| Editing Efficacy | KAMEL 1 token | Editing Efficacy100 | 4 |