Multiple-Choice QA on CLOTH
84.8AccuracyDirect-Joint
Evaluation Results
| Method | Links | |
|---|---|---|
| Direct-JointBackbone Model=Qwen-2.5-7B, Evaluation Protocol=0-shot2026.01 | 84.8 | |
| GTCABackbone Model=Qwen-2.5-7B, Evaluation Protocol=0-shot2026.01 | 83.98 | |
| LoRA-onlyBackbone Model=Qwen-2.5-7B, Evaluation Protocol=0-shot2026.01 | 83.71 | |
| GTCABackbone Model=Llama-3-8B, Evaluation Protocol=0-shot2026.01 | 82.74 | |
| LoRA-onlyBackbone Model=Llama-3-8B, Evaluation Protocol=0-shot2026.01 | 81.38 | |
| Direct-JointBackbone Model=Llama-3-8B, Evaluation Protocol=0-shot2026.01 | 80.46 | |
| BackboneBackbone Model=Qwen-2.5-7B, Evaluation Protocol=0-shot2026.01 | 74.6 | |
| BackboneBackbone Model=Llama-3-8B, Evaluation Protocol=0-shot2026.01 | 41.79 |