Multi-Hop Question Answering on HotpotQA (test val)
47.1Exact Match (EM)
DeepControl
Updated 17d ago
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
2026.02 | 47.1 | 47.9 | - | |
2026.07 | 42.9 | - | 55.3 | |
2026.07 | 41 | - | 52.9 | |
2026.07 | 40.4 | - | 52.2 | |
2026.02 | 40.2 | 41.1 | - | |
2026.07 | 40.1 | - | 51.6 | |
2026.07 | 39.7 | - | 51.5 | |
2026.07 | 39.6 | - | 50.8 | |
2026.07 | 39.1 | - | 50.6 | |
2026.07 | 39 | - | 50.6 | |
2026.07 | 38.7 | - | 49.7 | |
2026.07 | 38.4 | - | 49.4 | |
2026.07 | 38 | - | 48.6 | |
2026.07 | 37 | - | 47.3 | |
2026.07 | 36.9 | - | 48 | |
2026.07 | 35 | - | 45.9 | |
2026.02 | 18.3 | 18.1 | - | |
2026.02 | 14.9 | 13.4 | - |