Multi-hop Question Generation on HotpotQA
89.5Pairwise BLEUComposition
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| Composition2025.08 | 89.5 | 29 | 74.8 | |
| RAST2025.08 | 71.4 | 38.6 | 69.6 | |
| KCS2025.08 | 68.1 | 46.2 | 68 | |
| GPT-42025.08 | 15 | 79.8 | 91.8 |
| Method | Links | |||
|---|---|---|---|---|
| Composition2025.08 | 89.5 | 29 | 74.8 | |
| RAST2025.08 | 71.4 | 38.6 | 69.6 | |
| KCS2025.08 | 68.1 | 46.2 | 68 | |
| GPT-42025.08 | 15 | 79.8 | 91.8 |