Commonsense Reasoning on ARC Hard
42.41AccuracyBYOL-nya
Evaluation Results
| Method | Links | |
|---|---|---|
| BYOL-nyaParameters=12B, Pre-training Type=CPT2026.01 | 42.41 | |
| BYOL-nyaParameters=4B, Pre-training Type=CPT2026.01 | 40.61 | |
| Gemma-3Parameters=12B, Pre-training Type=PT2026.01 | 35.41 | |
| ApertusParameters=8B, Pre-training Type=25092026.01 | 27.73 | |
| BYOL-nyaParameters=1B, Pre-training Type=CPT2026.01 | 27.56 | |
| Gemma-3Parameters=4B, Pre-training Type=PT2026.01 | 27.13 | |
| Qwen-3Parameters=14B, Pre-training Type=Base2026.01 | 25.68 | |
| Llama-3.1Parameters=8B2026.01 | 24.06 | |
| Qwen-3Parameters=8B, Pre-training Type=Base2026.01 | 23.98 | |
| Gemma-3Parameters=1B, Pre-training Type=PT2026.01 | 23.72 | |
| Llama-3.2Parameters=1B2026.01 | 22.44 | |
| Qwen-3Parameters=1.7B, Pre-training Type=Base2026.01 | 21.93 |