Long-form Answer Generation on StoryQA
0.9203Spearman CorrelationMerge
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| MergeGranularity=Combined2026.04 | 0.9203 | 0.8923 | |
| WPAGranularity=Instance-level2026.04 | 0.8767 | 0.8467 | |
| Coarse 3-levelGranularity=Context-bound2026.04 | 0.8303 | 0.7832 | |
| PCPGranularity=Instance-level2026.04 | 0.8125 | 0.7602 | |
| ROUGE-LGranularity=Instance-level2026.04 | 0.6346 | 0.5897 | |
| Coarse 5-levelGranularity=Task-level2026.04 | 0.6075 | 0.5647 | |
| ChecklistGranularity=Instance-level2026.04 | 0.6061 | 0.5877 | |
| BLEUGranularity=Instance-level2026.04 | 0.5577 | 0.4872 |