Question Answering on 2WikiMultiHop
38.4EMVerbal-Calibrate
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| Verbal-CalibrateScope=global2026.04 | 38.4 | 42.9 | 48.2 | |
| Uncertain-CalibrateScope=local2026.04 | 36.2 | 39.6 | 59.2 | |
| DRAGIN2026.04 | 34.4 | 43.9 | 85.2 | |
| Ret-All2026.04 | 30 | 34.7 | 100 | |
| ADARAGUE2026.04 | 20.6 | 25.8 | 57.4 | |
| Base-UncTok2026.04 | 17.7 | 21.8 | 2.6 | |
| Base-Verbal2026.04 | 17 | 19.7 | 9.2 | |
| No-Ret2026.04 | 16.4 | 20.1 | — | |
| FLARE2026.04 | 15.2 | 22.6 | 99.4 | |
| SR-7BBackbone=7B2026.04 | 4.4 | 14.8 | 45.6 | |
| SR-13BBackbone=13B2026.04 | 3.4 | 16.6 | 30 |