NLG Meta-evaluation on CUS-QA en cs
0.73Kendall CorrelationLlama 4 Scout
Evaluation Results
| Method | Links | |
|---|---|---|
| Llama 4 ScoutShot=Few2026.03 | 0.73 | |
| Qwen 3 30BShot=Zero2026.03 | 0.725 | |
| Llama 4 ScoutShot=Zero2026.03 | 0.656 | |
| Qwen 3 30BShot=Few2026.03 | 0.651 | |
| Llama 3.3 70BShot=Few2026.03 | 0.619 | |
| Llama 3.3 70BShot=Zero2026.03 | 0.524 |