| Dataset Name | SOTA Method | Metric | Trend | ||
|---|---|---|---|---|---|
| MRCR | Score89.7 | 25 | 1mo ago | ||
| Long Context benchmark | AdaLoRA | Accuracy67.59 | 14 | 2mo ago | |
| NIAH 256K | Nemotron-Cascade-2-30B-A3B | Accuracy100 | 6 | 18d ago | |
| NIAH | Llama3.1 | Accuracy99.8 | 6 | 4mo ago | |
| AA-LCR | Nemotron-3-Puzzle-75B-A9B | Score56.6 | 5 | 18d ago | |
| RULER | Super | RULER Accuracy (256k Context)96.7 | 5 | 18d ago | |
| AA-LCR | Super | Accuracy57.3 | 4 | 18d ago | |
| RULER 128K | N-3-Super 120B-A12B-Base | Accuracy88.26 | 3 | 3mo ago | |
| NIAH 1M | Audex 30B-A3B | Score83.4 | 2 | 18d ago | |
| Leval | Score42.8 | 2 | 2mo ago | ||
| RULER | Score95 | 2 | 2mo ago | ||
| FRAMES | Kimi-Linear-48B-A3B-Instruct | Accuracy55.7 | 2 | 4mo ago | |
| RULER 1M | N-3-Super 120B-A12B-Base | Accuracy71 | 1 | 3mo ago | |
| RULER 512K | N-3-Super 120B-A12B-Base | Accuracy82.49 | 1 | 3mo ago |
