Multi-hop Question Answering on HotpotQA Wiki-18
44.7Exact MatchEsearch-instruct
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Esearch-instructModel Backbone=Qwen2.5-3b, Search Engine=Wiki-18, Model Variant=Instruct2025.10 | 44.7 | 58.7 | |
| Esearch-instructModel Backbone=Qwen2.5-7b, Search Engine=Wiki-18, Model Variant=Instruct2025.10 | 44.2 | 57.6 | |
| Esearch-baseModel Backbone=Qwen2.5-7b, Search Engine=Wiki-18, Model Variant=Base2025.10 | 43.4 | 56.4 | |
| Esearch-baseModel Backbone=Qwen2.5-3b, Search Engine=Wiki-18, Model Variant=Base2025.10 | 41.5 | 54.8 | |
| RAGModel Backbone=Qwen2.5-7b, Search Engine=Wiki-182025.10 | 31.7 | 37.5 | |
| RAGModel Backbone=Qwen2.5-3b, Search Engine=Wiki-182025.10 | 28.5 | 36.6 | |
| R1-instructModel Backbone=Qwen2.5-7b, Search Engine=Wiki-18, Model Variant=Instruct2025.10 | 25.4 | 31.4 | |
| R1-baseModel Backbone=Qwen2.5-3b, Search Engine=Wiki-18, Model Variant=Base2025.10 | 23.9 | 29.4 | |
| SFTModel Backbone=Qwen2.5-7b, Search Engine=Wiki-182025.10 | 23.3 | 28.7 | |
| R1-baseModel Backbone=Qwen2.5-7b, Search Engine=Wiki-18, Model Variant=Base2025.10 | 21.2 | 30.1 | |
| Search-o1Model Backbone=Qwen2.5-7b, Search Engine=Wiki-182025.10 | 20.6 | 25.7 | |
| Search-o1Model Backbone=Qwen2.5-3b, Search Engine=Wiki-182025.10 | 20.4 | 28.7 | |
| Direct InferenceModel Backbone=Qwen2.5-7b, Search Engine=Wiki-182025.10 | 20.1 | 24.8 | |
| SFTModel Backbone=Qwen2.5-3b, Search Engine=Wiki-182025.10 | 19.7 | 25.2 | |
| R1-instructModel Backbone=Qwen2.5-3b, Search Engine=Wiki-18, Model Variant=Instruct2025.10 | 19.4 | 27.9 | |
| Direct InferenceModel Backbone=Qwen2.5-3b, Search Engine=Wiki-182025.10 | 16.7 | 21.4 | |
| IRCoTModel Backbone=Qwen2.5-7b, Search Engine=Wiki-182025.10 | 12.1 | 20.6 | |
| CoTModel Backbone=Qwen2.5-7b, Search Engine=Wiki-182025.10 | 7.9 | 16.5 | |
| IRCoTModel Backbone=Qwen2.5-3b, Search Engine=Wiki-182025.10 | 7.7 | 13.5 | |
| CoTModel Backbone=Qwen2.5-3b, Search Engine=Wiki-182025.10 | 3.7 | 9.9 |