General Reasoning on Timely-Eval
82.9MATHDeepSeek-V3.2
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| DeepSeek-V3.22026.01 | 82.9 | 43.3 | 58.7 | |
| TimelyLM-8Bsize=8B2026.01 | 78 | 42.5 | 49.5 | |
| Qwen3-32Bsize=32B2026.01 | 75 | 45.7 | 35.5 | |
| GPT-5.1(medium)variant=medium2026.01 | 71.5 | 46.7 | 71 | |
| Qwen3-8Bsize=8B2026.01 | 71.2 | 40 | 37.5 | |
| Qwen3-14Bsize=14B2026.01 | 70.8 | 41.7 | 21 | |
| Gemini2.5-provariant=pro2026.01 | 63 | 37.5 | 59 |