Mathematical Reasoning on MATH500 Long context
89.12Accuracy0.7 × RoPE theta × 16 + 0.3 × 1M-Model
Evaluation results
| Method | Links | ||
|---|---|---|---|
| 0.7 × RoPE theta × 16 + 0.3 × 1M-ModelOperation=0.7 × RoPE theta × 16 + 0.3 × 1M-Model, recipe=LCTMerge2025.05 | 89.12 | 6,515 | |
| RoPE theta × 16Operation=RoPE theta × 162025.05 | 88.72 | 5,789 | |
| RoPE theta × 1Operation=RoPE theta × 12025.05 | 83.8 | 2,147 |