Long-form Answer Generation on ReviewSumm
0.4383Spearman CorrelationMerge
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Merge2026.04 | 0.4383 | 0.3891 | |
| WPA2026.04 | 0.427 | 0.3813 | |
| Coarse 5-level2026.04 | 0.3904 | 0.368 | |
| PCP2026.04 | 0.2357 | 0.2071 | |
| Checklist2026.04 | 0.2242 | 0.2108 | |
| Coarse 3-level2026.04 | 0.1607 | 0.149 | |
| BLEU2026.04 | -0.3619 | -0.3371 | |
| ROUGE-L2026.04 | -0.3993 | 0.3607 |