Multi-Hop Question Answering on Bamboogle (token-level F1)
65.2Token-Level F1Oracle
Evaluation Results
| Method | Links | |
|---|---|---|
| OracleMain agent backbone=Qwen2.5-7B-Instruct2026.04 | 65.2 | |
| COSEARCHMain agent backbone=Qwen2.5-7B-Instruct2026.04 | 60.1 | |
| OracleMain agent backbone=Qwen2.5-3B-Instruct2026.04 | 58.9 | |
| Fixed RankerMain agent backbone=Qwen2.5-7B-Instruct2026.04 | 58.8 | |
| Retrieval OnlyMain agent backbone=Qwen2.5-7B-Instruct2026.04 | 57.4 | |
| ZeroSearchMain agent backbone=Qwen2.5-7B-Instruct2026.04 | 57.2 | |
| Search-R1Main agent backbone=Qwen2.5-7B-Instruct2026.04 | 55.8 | |
| COSEARCHMain agent backbone=Qwen2.5-3B-Instruct2026.04 | 45.7 | |
| Fixed RankerMain agent backbone=Qwen2.5-3B-Instruct2026.04 | 44.8 | |
| Retrieval OnlyMain agent backbone=Qwen2.5-3B-Instruct2026.04 | 43.6 | |
| ZeroSearchMain agent backbone=Qwen2.5-3B-Instruct2026.04 | 43.1 | |
| Search-o1Main agent backbone=Qwen2.5-7B-Instruct2026.04 | 42.2 | |
| Search-R1Main agent backbone=Qwen2.5-3B-Instruct2026.04 | 41.4 | |
| Direct InferenceMain agent backbone=Qwen2.5-7B-Instruct2026.04 | 37.2 | |
| CoTMain agent backbone=Qwen2.5-7B-Instruct2026.04 | 31 | |
| RAGMain agent backbone=Qwen2.5-7B-Instruct2026.04 | 29.6 | |
| Search-o1Main agent backbone=Qwen2.5-3B-Instruct2026.04 | 21.6 | |
| CoTMain agent backbone=Qwen2.5-3B-Instruct2026.04 | 10 | |
| Direct InferenceMain agent backbone=Qwen2.5-3B-Instruct2026.04 | 6.8 | |
| RAGMain agent backbone=Qwen2.5-3B-Instruct2026.04 | 5.8 |