Long-context understanding on RULER (4K to 64K Contexts)
96.56Accuracy (4K)
Sketch&Walk
Updated 4d ago
Evaluation Results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
2026.02 | 96.56 | 94.06 | 94.06 | 89.63 | 82.5 | 91.35 | |
2026.02 | 95.31 | 95.31 | 94.06 | 89.69 | 82.81 | 91.44 | |
2026.02 | 95.31 | 92.5 | 93.75 | 88.13 | 80.31 | 90 | |
2026.02 | 92.65 | 95 | 93.12 | 86.87 | 82.5 | 90.03 | |
2026.02 | 74.38 | 68.44 | 65.31 | 66.88 | 66.56 | 68.31 | |
2026.02 | 74.06 | 63.75 | 60.63 | 61.88 | 55.94 | 63.25 | |
2026.02 | 74.06 | 68.44 | 66.06 | 65.62 | 63.75 | 67.59 | |
2026.02 | 73.44 | 69.06 | 64.69 | 64.38 | 62.71 | 66.86 |