NLG Evaluation on USR
0.82Spearman CorrelationGPT-4o
Evaluation Results
| Method | Links | |
|---|---|---|
| GPT-4oPrompting=Few-shot2026.02 | 0.82 | |
| GPT-4oPrompting=Zero-shot2026.02 | 0.81 | |
| QwenPrompting=Zero-shot2026.02 | 0.81 | |
| QwenPrompting=Few-shot2026.02 | 0.8 | |
| GPT-4o-miniPrompting=Zero-shot2026.02 | 0.79 | |
| LlamaPrompting=Few-shot2026.02 | 0.79 | |
| MixtralPrompting=Zero-shot2026.02 | 0.79 | |
| MixtralPrompting=Few-shot2026.02 | 0.79 | |
| GPT-4o-miniPrompting=Few-shot2026.02 | 0.78 | |
| LlamaPrompting=Zero-shot2026.02 | 0.77 |