Mathematical Reasoning on AIME24 (EM)
93.33Exact Match (EM)DeepSeek-R1-0528
Evaluation Results
| Method | Links | |
|---|---|---|
| DeepSeek-R1-0528Maximum sequence length=128k, KV cache length=Full2025.11 | 93.33 | |
| SnapStreamMaximum sequence length=128k, KV cache length=32k2025.11 | 90 | |
| DeepSeek-R1-0528Maximum sequence length=32k, KV cache length=32k2025.11 | 86.67 |