Monolingual Question Answering on 2WIKI
61.13fEMCroSearch-R1
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| CroSearch-R1Backbone=Qwen2.5-7B-Instruct, Paradigm=Post-training2026.04 | 61.13 | 73.64 | |
| Search-R1Backbone=Qwen2.5-7B-Instruct, Paradigm=Post-training2026.04 | 56.05 | 66.08 | |
| CroSearch-R1Backbone=Qwen2.5-3B-Instruct, Paradigm=Post-training2026.04 | 55.26 | 70.42 | |
| Search-R1Backbone=Qwen2.5-3B-Instruct, Paradigm=Post-training2026.04 | 52.93 | 65.76 | |
| SFTBackbone=Qwen2.5-7B-Instruct, Paradigm=Post-training2026.04 | 39.8 | 46.25 | |
| RAGBackbone=Qwen2.5-7B-Instruct, Paradigm=Training-free2026.04 | 39.7 | 46.69 | |
| SFTBackbone=Qwen2.5-3B-Instruct, Paradigm=Post-training2026.04 | 38.9 | 45.9 | |
| RAGBackbone=Qwen2.5-3B-Instruct, Paradigm=Training-free2026.04 | 38.6 | 45.9 | |
| Direct InferenceBackbone=Qwen2.5-3B-Instruct, Paradigm=Training-free2026.04 | 36.6 | 44.79 | |
| IRCoTBackbone=Qwen2.5-3B-Instruct, Paradigm=Training-free2026.04 | 34.17 | 44.58 | |
| Search-o1Backbone=Qwen2.5-7B-Instruct, Paradigm=Training-free2026.04 | 33.92 | 46.88 | |
| Direct InferenceBackbone=Qwen2.5-7B-Instruct, Paradigm=Training-free2026.04 | 33.8 | 41.59 | |
| IRCoTBackbone=Qwen2.5-7B-Instruct, Paradigm=Training-free2026.04 | 33.34 | 44.85 | |
| Search-o1Backbone=Qwen2.5-3B-Instruct, Paradigm=Training-free2026.04 | 33.12 | 45.65 |