Multi-hop Question Answering on 2Wiki 18
43.6Exact Match (EM)Esearch-base
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Esearch-baseModel Backbone=Qwen2.5-7b, Search Engine=Wiki-18, Model Variant=Base2025.10 | 43.6 | 51.3 | |
| Esearch-baseModel Backbone=Qwen2.5-3b, Search Engine=Wiki-18, Model Variant=Base2025.10 | 42.8 | 49.9 | |
| Esearch-instructModel Backbone=Qwen2.5-7b, Search Engine=Wiki-18, Model Variant=Instruct2025.10 | 41.9 | 49.4 | |
| Esearch-instructModel Backbone=Qwen2.5-3b, Search Engine=Wiki-18, Model Variant=Instruct2025.10 | 41.5 | 50 | |
| R1-instructModel Backbone=Qwen2.5-7b, Search Engine=Wiki-18, Model Variant=Instruct2025.10 | 30.4 | 36.1 | |
| SFTModel Backbone=Qwen2.5-7b, Search Engine=Wiki-182025.10 | 27.7 | 32.8 | |
| Direct InferenceModel Backbone=Qwen2.5-3b, Search Engine=Wiki-182025.10 | 26.3 | 30.8 | |
| R1-baseModel Backbone=Qwen2.5-3b, Search Engine=Wiki-18, Model Variant=Base2025.10 | 26.2 | 31.7 | |
| R1-instructModel Backbone=Qwen2.5-3b, Search Engine=Wiki-18, Model Variant=Instruct2025.10 | 23.9 | 32.7 | |
| Direct InferenceModel Backbone=Qwen2.5-7b, Search Engine=Wiki-182025.10 | 23.8 | 31.9 | |
| Search-o1Model Backbone=Qwen2.5-3b, Search Engine=Wiki-182025.10 | 23 | 29.3 | |
| R1-baseModel Backbone=Qwen2.5-7b, Search Engine=Wiki-18, Model Variant=Base2025.10 | 22.9 | 31.5 | |
| RAGModel Backbone=Qwen2.5-7b, Search Engine=Wiki-182025.10 | 22.1 | 30.7 | |
| RAGModel Backbone=Qwen2.5-3b, Search Engine=Wiki-182025.10 | 19.2 | 27.1 | |
| Search-o1Model Backbone=Qwen2.5-7b, Search Engine=Wiki-182025.10 | 18.9 | 24.6 | |
| SFTModel Backbone=Qwen2.5-3b, Search Engine=Wiki-182025.10 | 15.8 | 24.3 | |
| IRCoTModel Backbone=Qwen2.5-3b, Search Engine=Wiki-182025.10 | 13.7 | 19.7 | |
| IRCoTModel Backbone=Qwen2.5-7b, Search Engine=Wiki-182025.10 | 13.3 | 21.8 | |
| CoTModel Backbone=Qwen2.5-7b, Search Engine=Wiki-182025.10 | 12.7 | 18.4 | |
| CoTModel Backbone=Qwen2.5-3b, Search Engine=Wiki-182025.10 | 1.6 | 9.4 |