Question Answering on PopQA (Accuracy (Acc))
70.7Accuracy (Acc)Qwen3-32B+Hit-RAG
Evaluation Results
| Method | Links | |
|---|---|---|
| Qwen3-32B+Hit-RAGFramework=Hit-RAG2026.03 | 70.7 | |
| Llama3.1-70B+R-InsRetrieval=Yes2026.03 | 70.4 | |
| Llama3.1-8B+R-InsRetrieval=Yes2026.03 | 68.4 | |
| Llama3.1-70B+Hit-RAGFramework=Hit-RAG2026.03 | 66.7 | |
| RankRAG-70BRetrieval=Yes2026.03 | 65.4 | |
| RankRAG-8BRetrieval=Yes2026.03 | 64.1 | |
| Llama3.1-8B+Hit-RAGFramework=Hit-RAG2026.03 | 63.6 | |
| Qwen3-8B+Hit-RAGFramework=Hit-RAG2026.03 | 63.1 | |
| Qwen3-4B+Hit-RAGFramework=Hit-RAG2026.03 | 60.8 | |
| RAS-7BRetrieval=Yes2026.03 | 58.3 | |
| RAS-8BRetrieval=Yes2026.03 | 57.7 | |
| Speculative RAG-7BRetrieval=Yes2026.03 | 57.5 | |
| Self-Reasoning-13BRetrieval=Yes2026.03 | 57.3 | |
| Self-RAG-13BRetrieval=Yes2026.03 | 55.8 | |
| Self-RAG-7BRetrieval=Yes2026.03 | 54.9 | |
| Self-Reasoning-7BRetrieval=Yes2026.03 | 54.2 | |
| Ret-ChatGPTRetrieval=None2026.03 | 50.8 | |
| Alpaca-13BRetrieval=Yes2026.03 | 46.1 | |
| Llama3.1-70BRetrieval=None2026.03 | 31.1 | |
| Sonnet-3.5Retrieval=None2026.03 | 30.2 | |
| ChatGPTRetrieval=None2026.03 | 29.3 | |
| Llama3.1-8BRetrieval=None2026.03 | 28.8 | |
| Qwen3-32BRetrieval=None2026.03 | 25.1 | |
| Alpaca-13BRetrieval=None2026.03 | 24.4 | |
| Qwen3-8BRetrieval=None2026.03 | 20.6 | |
| Qwen3-4BRetrieval=None2026.03 | 16.5 |