Multiple choice question answering on BT (test)
39.6AccuracyBase
Evaluation Results
| Method | Links | |
|---|---|---|
| BaseBackbone=QWEN2-7B2026.05 | 39.6 | |
| PPT (Mean)Backbone=QWEN2-7B2026.05 | 39.4 | |
| PPT (Distribution)Backbone=LLAMA3-8B2026.05 | 39.3 | |
| PPT (Distribution)Backbone=QWEN2-7B2026.05 | 38.9 | |
| BaseBackbone=LLAMA3-8B2026.05 | 37.8 | |
| PPT (Mean)Backbone=LLAMA3-8B2026.05 | 36.8 | |
| VerbalizedBackbone=LLAMA3-8B2026.05 | 35.3 |