Question Answering on Pub
84.1AccuracyLlama3.1-70B+Hit-RAG
Evaluation Results
| Method | Links | |
|---|---|---|
| Llama3.1-70B+Hit-RAGFramework=Hit-RAG2026.03 | 84.1 | |
| Sonnet-3.5Retrieval=None2026.03 | 83.7 | |
| Llama3.1-70B+R-InsRetrieval=Yes2026.03 | 82.8 | |
| Qwen3-32B+Hit-RAGFramework=Hit-RAG2026.03 | 82.4 | |
| Llama3.1-8B+Hit-RAGFramework=Hit-RAG2026.03 | 81.5 | |
| Qwen3-8B+Hit-RAGFramework=Hit-RAG2026.03 | 80.1 | |
| RAS-8BRetrieval=Yes2026.03 | 77.6 | |
| Llama3.1-8B+R-InsRetrieval=Yes2026.03 | 77.2 | |
| Speculative RAG-7BRetrieval=Yes2026.03 | 75.7 | |
| RAS-7BRetrieval=Yes2026.03 | 74.7 | |
| Self-RAG-13BRetrieval=Yes2026.03 | 74.5 | |
| Qwen3-4B+Hit-RAGFramework=Hit-RAG2026.03 | 73.3 | |
| Self-RAG-7BRetrieval=Yes2026.03 | 72.4 | |
| ChatGPTRetrieval=None2026.03 | 70.1 | |
| Alpaca-13BRetrieval=None2026.03 | 55.5 | |
| Ret-ChatGPTRetrieval=None2026.03 | 54.7 | |
| Alpaca-13BRetrieval=Yes2026.03 | 51.1 | |
| Qwen3-32BRetrieval=None2026.03 | 48.7 | |
| Llama3.1-70BRetrieval=None2026.03 | 43.5 | |
| Qwen3-8BRetrieval=None2026.03 | 40.9 | |
| Qwen3-4BRetrieval=None2026.03 | 38.3 | |
| Llama3.1-8BRetrieval=None2026.03 | 35.8 |