Multiple Choice Question Answering on MMLU-Pro Health discipline
9.38Mean EMRQwen-2-7B-Instruct
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| Qwen-2-7B-InstructOutlier Detection=SConU (current paper), alpha=0.12025.04 | 9.38 | 2.83 | 9.6 | |
| LLaMA-3.1-8B-InstructOutlier Detection=ConU (basic framework), alpha=0.12025.04 | 9.61 | 2.73 | 9.6 | |
| LLaMA-3.1-8B-InstructOutlier Detection=SConU (current paper), alpha=0.12025.04 | 9.75 | 2.14 | 9.6 | |
| Qwen-2-7B-InstructOutlier Detection=ConU (basic framework), alpha=0.12025.04 | 10.19 | 2.85 | 10.8 | |
| Qwen-2-7B-InstructOutlier Detection=SConU (current paper), alpha=0.22025.04 | 19.43 | 3.62 | 19.2 | |
| Qwen-2-7B-InstructOutlier Detection=ConU (basic framework), alpha=0.22025.04 | 19.77 | 3.72 | 19.6 | |
| Qwen-2-7B-InstructOutlier Detection=SConU (current paper), alpha=0.32025.04 | 29.72 | 4.23 | 29.2 | |
| Qwen-2-7B-InstructOutlier Detection=ConU (basic framework), alpha=0.32025.04 | 30.01 | 4.2 | 29.6 | |
| Qwen-2-7B-InstructOutlier Detection=SConU (current paper), alpha=0.42025.04 | 39.57 | 4.25 | 39.6 | |
| Qwen-2-7B-InstructOutlier Detection=ConU (basic framework), alpha=0.42025.04 | 40.35 | 4.34 | 41.2 |