Long-form Answer Generation on MultiDocQA
0.392Spearman CorrelationMerge
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Merge2026.04 | 0.392 | 0.3647 | |
| WPA2026.04 | 0.3381 | 0.3059 | |
| Coarse 5-level2026.04 | 0.3253 | 0.3055 | |
| Coarse 3-level2026.04 | 0.3199 | 0.3004 | |
| PCP2026.04 | 0.2576 | 0.2455 | |
| Checklist2026.04 | 0.2273 | 0.2132 | |
| ROUGE-L2026.04 | 0.0751 | 0.0591 | |
| BLEU2026.04 | 0.0691 | 0.051 |