General Reasoning on MMLU (0-shot test)
64.75Pass@1 AccuracyHEX
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| HEXGen. Len.=256, Training Necessity=Training-free, Setting=0-shot, reimplementation=true2026.02 | 64.75 | 3,200 | |
| HEXGen. Len.=128, Training Necessity=Training-free, Setting=0-shot, reimplementation=true2026.02 | 64.28 | 1,600 | |
| Majority VotingGen. Len.=256, Training Necessity=Training-free, Setting=0-shot2026.02 | 63.11 | 640 | |
| dVOTINGGen. Len.=128, Training Necessity=Training-free, Setting=0-shot2026.02 | 62.87 | 222.1 | |
| dVOTINGGen. Len.=256, Training Necessity=Training-free, Setting=0-shot2026.02 | 62.87 | 391.4 | |
| Majority VotingGen. Len.=128, Training Necessity=Training-free, Setting=0-shot2026.02 | 62.27 | 320 | |
| LLaDA-8B-InstructGen. Len.=256, Training Necessity=Pre-training, Setting=0-shot2026.02 | 58.04 | 256 | |
| LLaDA-8B-InstructGen. Len.=128, Training Necessity=Pre-training, Setting=0-shot2026.02 | 57.13 | 128 |