Multi-hop Question Answering on Bamboogle (EM)
57.6EMMemSearcher
Evaluation Results
| Method | Links | |
|---|---|---|
| MemSearcherParameter Scale=14B2025.11 | 57.6 | |
| ReSearchParameter Scale=32B2025.11 | 56.8 | |
| Search-R1-baseParameter Scale=14B2025.11 | 52.8 | |
| MemSearcherParameter Scale=7B2025.11 | 48.8 | |
| Search-R1-instructParameter Scale=14B2025.11 | 48 | |
| Search-R1-baseParameter Scale=7B2025.11 | 43.2 | |
| MemSearcherParameter Scale=3B2025.11 | 42.4 | |
| ReSearchParameter Scale=7B2025.11 | 42.4 | |
| ReSearchParameter Scale=3B2025.11 | 37.6 | |
| Search-R1-instructParameter Scale=7B2025.11 | 36.8 | |
| R1-SearcherParameter Scale=7B, Search Environment=Google Web Search2025.11 | 36.8 | |
| AutoRefine-baseParameter Scale=3B2025.11 | 34.4 | |
| O2-SearcherParameter Scale=3B2025.11 | 34.4 | |
| Search-o1Parameter Scale=14B2025.11 | 34.4 | |
| AutoRefine-instructParameter Scale=3B2025.11 | 33.6 | |
| Search-o1Parameter Scale=7B2025.11 | 30.4 | |
| ZeroSearchParameter Scale=7B, Search Environment=Google Web Search2025.11 | 27.8 | |
| Search-R1-instructParameter Scale=3B2025.11 | 26.4 | |
| Search-o1Parameter Scale=3B2025.11 | 22.4 | |
| ZeroSearchParameter Scale=3B, Search Environment=Google Web Search2025.11 | 11.1 | |
| Search-R1-baseParameter Scale=3B2025.11 | 8.8 |