Commonsense Reasoning on XCSQA
64.6CLCall ScoreQwen2.5-14B
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| Qwen2.5-14BModel=Qwen2.5-14B2026.03 | 64.6 | 87 | 56.9 | |
| Aya-Expanse-8BModel=Aya-Expanse-8B2026.03 | 62.6 | 78 | 54.4 | |
| Qwen3-14BModel=Qwen3-14B2026.03 | 61.9 | 77.6 | 54 | |
| Llama3.1-8BModel=Llama3.1-8B2026.03 | 60.2 | 67.5 | 47.7 | |
| Gemma3-12B-ptModel=Gemma3-12B-pt2026.03 | 58.3 | 66 | 47.2 | |
| DCOBase Model=Llama3.1-8B2026.03 | 9.1 | 0.2 | 1.3 | |
| DCOBase Model=Qwen3-14B2026.03 | 7.1 | 1.1 | 3.8 | |
| DCOBase Model=Qwen2.5-14B2026.03 | 6.8 | -2.5 | 4.7 | |
| DCOBase Model=Aya-Expanse-8B2026.03 | 6.4 | 0.6 | 3.7 | |
| DCOBase Model=Gemma3-12B-pt2026.03 | 4.6 | 0.1 | 3.6 |