Answer-position balance evaluation on MCQ
0.061MCQ TVGPT-OSS-20B
Evaluation Results
| Method | Links | |
|---|---|---|
| GPT-OSS-20BCalibration=Soft2026.05 | 0.061 | |
| Llama-3.2-3B-itCalibration=Base2026.05 | 0.066 | |
| GPT-OSS-20BCalibration=Hard2026.05 | 0.072 | |
| Qwen3-4BCalibration=Soft2026.05 | 0.116 | |
| Qwen3-4BCalibration=Hard2026.05 | 0.156 | |
| Llama-3.2-1B-itCalibration=Base2026.05 | 0.159 | |
| GPT-OSS-20BCalibration=Base2026.05 | 0.166 | |
| Gemma-3-4B-itCalibration=Soft2026.05 | 0.178 | |
| Qwen3-14BCalibration=Soft2026.05 | 0.184 | |
| Llama-3.2-3B-itCalibration=Soft2026.05 | 0.187 | |
| Llama-3.2-3B-itCalibration=Hard2026.05 | 0.189 | |
| Qwen3-14BCalibration=Hard2026.05 | 0.19 | |
| Qwen3-4BCalibration=Base2026.05 | 0.192 | |
| Llama-3.2-1B-itCalibration=Hard2026.05 | 0.199 | |
| Gemma-3-1B-itCalibration=Hard2026.05 | 0.203 | |
| Gemma-3-1B-itCalibration=Soft2026.05 | 0.205 | |
| Gemma-3-1B-itCalibration=Base2026.05 | 0.209 | |
| Qwen3-8BCalibration=Hard2026.05 | 0.221 | |
| Gemma-3-27B-itCalibration=Hard2026.05 | 0.238 | |
| Qwen3-8BCalibration=Base2026.05 | 0.24 | |
| Qwen3-8BCalibration=Soft2026.05 | 0.24 | |
| Gemma-3-4B-itCalibration=Base2026.05 | 0.24 | |
| Qwen3-1.7BCalibration=Soft2026.05 | 0.249 | |
| Gemma-3-4B-itCalibration=Hard2026.05 | 0.256 | |
| Qwen3-14BCalibration=Base2026.05 | 0.284 | |
| Gemma-3-27B-itCalibration=Base2026.05 | 0.284 | |
| Qwen3-1.7BCalibration=Hard2026.05 | 0.288 | |
| Gemma-3-12B-itCalibration=Soft2026.05 | 0.289 | |
| Qwen3-1.7BCalibration=Base2026.05 | 0.313 | |
| Gemma-3-12B-itCalibration=Hard2026.05 | 0.313 | |
| Gemma-3-27B-itCalibration=Soft2026.05 | 0.315 | |
| Gemma-3-12B-itCalibration=Base2026.05 | 0.335 | |
| Qwen3-0.6BCalibration=Base2026.05 | 0.374 | |
| Qwen3-0.6BCalibration=Hard2026.05 | 0.398 |