Long-form Answer Generation on ContractQA
0.564Spearman CorrelationCoarse 5-level
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Coarse 5-level2026.04 | 0.564 | 0.5273 | |
| Merge2026.04 | 0.5268 | 0.4785 | |
| PCP2026.04 | 0.4719 | 0.4364 | |
| WPA2026.04 | 0.4696 | 0.4337 | |
| Coarse 3-level2026.04 | 0.4604 | 0.4293 | |
| Checklist2026.04 | 0.416 | 0.39 | |
| BLEU2026.04 | 0.2619 | 0.2381 | |
| ROUGE-L2026.04 | 0.2202 | 0.2103 |