Question Answering on HotpotQA (Mean/Std/p-value)
38.8Mean PerformanceReSeek
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| ReSeekModel Size=7B2025.10 | 38.8 | 0.007 | 0.003 | |
| ZeroSearchModel Size=7B2025.10 | 34.6 | 0.005 | — | |
| ReSeekModel Size=3B2025.10 | 32.4 | 0.008 | 0.028 | |
| ZeroSearchModel Size=3B2025.10 | 27.4 | 0.006 | — |