Machine Reading Comprehension on BELEBELE Korean
68.8Accuracy (Target)CLO
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| CLOModel=Qwen2.5-3B, Zero-shot=true2025.05 | 68.8 | 82 | |
| SFTModel=Qwen2.5-3B, Zero-shot=true2025.05 | 67.7 | 80.9 | |
| CLOModel=Llama-3-8B, Zero-shot=true2025.05 | 57.7 | 73.4 | |
| CLOModel=Mistral-7B-v0.1, Zero-shot=true2025.05 | 48 | 58.8 | |
| SFTModel=Llama-3-8B, Zero-shot=true2025.05 | 46.7 | 64.9 | |
| CLOModel=Llama-2-13B, Zero-shot=true2025.05 | 38.7 | 57.6 | |
| SFTModel=Llama-2-13B, Zero-shot=true2025.05 | 37.9 | 53.6 | |
| CLOModel=Llama-2-7B, Zero-shot=true2025.05 | 30.1 | 37.9 | |
| SFTModel=Mistral-7B-v0.1, Zero-shot=true2025.05 | 29.2 | 70.6 | |
| SFTModel=Llama-2-7B, Zero-shot=true2025.05 | 27.4 | 37.1 |