NLG Meta-evaluation on CUS-QA orig. (sk)
0.788Kendall CorrelationQwen 3 30B
Evaluation Results
| Method | Links | |
|---|---|---|
| Qwen 3 30BShot=Zero2026.03 | 0.788 | |
| Llama 4 ScoutShot=Few2026.03 | 0.735 | |
| Llama 3.3 70BShot=Few2026.03 | 0.73 | |
| Qwen 3 30BShot=Few2026.03 | 0.709 | |
| Llama 4 ScoutShot=Zero2026.03 | 0.704 | |
| Llama 3.3 70BShot=Zero2026.03 | 0.656 |