Open-domain Question Answering on Bamboogle out-of-domain
71.7F1Self-Ask
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Self-AskBackbone=GPT-4.1-mini, Type=Advanced RAG2025.08 | 71.7 | 75.2 | |
| R1-SearcherBackbone=LLaMA-3.1-8B, Type=RAG-RL2025.08 | 64.7 | 62.4 | |
| UR2Backbone=Qwen-2.5-7B, Type=Our Implementations2025.08 | 64.5 | 62.4 | |
| Search-o1Backbone=GPT-4.1-mini, Type=Advanced RAG2025.08 | 63.7 | 71.2 | |
| Vanilla RLBackbone=Qwen-2.5-7B, Type=Our Implementations2025.08 | 63.3 | 63.2 | |
| R1-Searcher++Backbone=Qwen-2.5-7B, Type=RAG-RL2025.08 | 60.8 | 59.2 | |
| UR2Backbone=LLaMA-3.1-8B, Type=Our Implementations2025.08 | 60.7 | 63.2 | |
| Vanilla RLBackbone=LLaMA-3.1-8B, Type=Our Implementations2025.08 | 60.6 | 63.2 | |
| RATBackbone=GPT-4.1-mini, Type=Advanced RAG2025.08 | 60.3 | 62.4 | |
| CoTBackbone=GPT-4.1-mini, Type=Vanilla Methods2025.08 | 59.2 | 76 | |
| Vanilla RLBackbone=Qwen-2.5-3B, Type=Our Implementations2025.08 | 59 | 57.6 | |
| Search-R1Backbone=Qwen-2.5-7B, Type=RAG-RL2025.08 | 58.9 | 56.8 | |
| UR2Backbone=Qwen-2.5-3B, Type=Our Implementations2025.08 | 57.8 | 58.4 | |
| R1-SearcherBackbone=Qwen-2.5-7B, Type=RAG-RL2025.08 | 56.5 | 53.6 | |
| Search-R1Backbone=Qwen-2.5-3B, Type=RAG-RL2025.08 | 48.3 | 48 | |
| Standard RAGBackbone=GPT-4.1-mini, Type=Vanilla Methods2025.08 | 46.4 | 51.2 | |
| CoTBackbone=LLaMA-3.1-8B, Type=Vanilla Methods2025.08 | 43 | 42.4 | |
| CoTBackbone=Qwen-2.5-7B, Type=Vanilla Methods2025.08 | 41.3 | 43.2 | |
| Standard RAGBackbone=Qwen-2.5-7B, Type=Vanilla Methods2025.08 | 38.9 | 40 | |
| ZeroSearchBackbone=Qwen-2.5-7B, Type=RAG-RL2025.08 | 35.8 | 38.4 | |
| Zero-SearchBackbone=Qwen-2.5-3B, Type=RAG-RL2025.08 | 32.4 | 31.2 | |
| CoTBackbone=Qwen-2.5-3B, Type=Vanilla Methods2025.08 | 31.2 | 33.6 | |
| Standard RAGBackbone=LLaMA-3.1-8B, Type=Vanilla Methods2025.08 | 26.5 | 28 | |
| Standard RAGBackbone=Qwen-2.5-3B, Type=Vanilla Methods2025.08 | 26.1 | 27.2 |