Single-hop Question Answering on NQ (Natural Questions) (test)
45.3AccuracyInfoReasoner-3B
Evaluation Results
| Method | Links | |
|---|---|---|
| InfoReasoner-3BRetrieval=true, Model Scale=3B2026.01 | 45.3 | |
| InfoReasoner-7BRetrieval=true, Model Scale=7B2026.01 | 44.7 | |
| AutoRefine-7B-BaseRetrieval=true, Model Scale=7B2026.01 | 43.9 | |
| Search-R1-7B-InstructRetrieval=true, Model Scale=7B2026.01 | 40.7 | |
| Search-R1-3B-InstructRetrieval=true, Model Scale=3B2026.01 | 40.5 | |
| Search-R1-3B-BaseRetrieval=true, Model Scale=3B2026.01 | 39.4 | |
| Search-R1-7B-BaseRetrieval=true, Model Scale=7B2026.01 | 39.1 | |
| InForge-3BRetrieval=true, Model Scale=3B2026.01 | 38.6 | |
| AutoRefine-3B-InstructRetrieval=true, Model Scale=3B2026.01 | 38.3 | |
| RAGRetrieval=true, Model Scale=3B2026.01 | 34.8 | |
| ReSearch-7B-InstructRetrieval=true, Model Scale=7B2026.01 | 33.3 | |
| ReasonRAG-7BRetrieval=true, Model Scale=7B2026.01 | 29.4 | |
| AutoRefine-3B-BaseRetrieval=true, Model Scale=3B2026.01 | 28.1 | |
| ReSearch-7B-BaseRetrieval=true, Model Scale=7B2026.01 | 26.8 | |
| SFTRetrieval=false, Model Scale=3B2026.01 | 24.9 | |
| Search-o1Retrieval=true, Model Scale=3B2026.01 | 23.8 | |
| R1-BaseRetrieval=false, Model Scale=3B2026.01 | 22.6 | |
| R1-InstructRetrieval=false, Model Scale=3B2026.01 | 21 | |
| IRCoTRetrieval=true, Model Scale=3B2026.01 | 11.1 | |
| Direct GenerationRetrieval=false, Model Scale=3B2026.01 | 10.6 | |
| CoTRetrieval=false, Model Scale=3B2026.01 | 4.8 |