| Benchmark | Model | Rank | Results |
|---|---|---|---|
| mathematical-reasoning-on-math-500 | Qwen2.5-32B-Instruct + segment SFT (s1K-1.1, Prefix) | #30 | Accuracy: 95 |
| mathematical-reasoning-on-aime24 | Qwen2.5-32B-Instruct + segment SFT (s1K-1.1, Both endpoints) | #51 | Acc: 68.89 |
| common-sense-reasoning-on-winogrande | Qwen3-4B-Base + E-SFT (OpenThoughts3-100K) | #75 | Accuracy: 70.64 |
| sentence-completion-on-hellaswag | Qwen3-4B-Base + E-SFT (OpenThoughts3-100K) | #166 | Accuracy: 54.86 |