Simple Question Answering on PopQA (test)
65.24RAG F1HippoRAG2 + DeepRefine-8B
Evaluation Results
| Method | Links | |
|---|---|---|
| HippoRAG2 + DeepRefine-8BConstructor=AR1, Refinement Model=DeepRefine-8B2026.05 | 65.24 | |
| HippoRAG2Constructor=AR12026.05 | 64.69 | |
| HippoRAG + DeepRefine-8BConstructor=AR1, Refinement Model=DeepRefine-8B2026.05 | 64.63 | |
| HippoRAGConstructor=AR12026.05 | 64.43 | |
| HippoRAG2 + DeepRefine-4BConstructor=AR1, Refinement Model=DeepRefine-4B2026.05 | 63.91 | |
| HippoRAG + DeepRefine-4BConstructor=AR1, Refinement Model=DeepRefine-4B2026.05 | 63.87 | |
| HippoRAG + DeepRefine-4BConstructor=Naive, Refinement Model=DeepRefine-4B2026.05 | 63.73 | |
| HippoRAGConstructor=Naive2026.05 | 63.56 | |
| HippoRAG2 + DeepRefine-8BConstructor=Naive, Refinement Model=DeepRefine-8B2026.05 | 63.4 | |
| HippoRAG2Constructor=Naive2026.05 | 63.08 | |
| HippoRAG + DeepRefine-8BConstructor=Naive, Refinement Model=DeepRefine-8B2026.05 | 62.96 | |
| HippoRAG2 + DeepRefine-4BConstructor=Naive, Refinement Model=DeepRefine-4B2026.05 | 62.81 | |
| ToG + DeepRefine-8BConstructor=AR1, Refinement Model=DeepRefine-8B2026.05 | 61.2 | |
| ToG + DeepRefine-4BConstructor=AR1, Refinement Model=DeepRefine-4B2026.05 | 60.5 | |
| ToGConstructor=AR12026.05 | 60.34 | |
| Subgraph + DeepRefine-4BConstructor=AR1, Refinement Model=DeepRefine-4B2026.05 | 59.07 | |
| Subgraph + DeepRefine-8BConstructor=AR1, Refinement Model=DeepRefine-8B2026.05 | 58.99 | |
| SubgraphConstructor=AR12026.05 | 58.4 | |
| Subgraph + DeepRefine-8BConstructor=Naive, Refinement Model=DeepRefine-8B2026.05 | 56.29 | |
| ToG + DeepRefine-4BConstructor=Naive, Refinement Model=DeepRefine-4B2026.05 | 56.1 | |
| Subgraph + DeepRefine-4BConstructor=Naive, Refinement Model=DeepRefine-4B2026.05 | 56 | |
| ToG + DeepRefine-8BConstructor=Naive, Refinement Model=DeepRefine-8B2026.05 | 55.66 | |
| ToGConstructor=Naive2026.05 | 54.92 | |
| SubgraphConstructor=Naive2026.05 | 54.48 | |
| Subgraph + DeepRefine-4BConstructor=Graphify (LLM-Wiki), Refinement Model=DeepRefine-4B2026.05 | 15.86 | |
| Subgraph + DeepRefine-8BConstructor=Graphify (LLM-Wiki), Refinement Model=DeepRefine-8B2026.05 | 15.78 | |
| SubgraphConstructor=Graphify (LLM-Wiki)2026.05 | 11.38 | |
| ToG + DeepRefine-4BConstructor=Graphify (LLM-Wiki), Refinement Model=DeepRefine-4B2026.05 | 10.88 | |
| ToG + DeepRefine-8BConstructor=Graphify (LLM-Wiki), Refinement Model=DeepRefine-8B2026.05 | 10.61 | |
| ToGConstructor=Graphify (LLM-Wiki)2026.05 | 8.97 | |
| HippoRAG2 + DeepRefine-8BConstructor=Graphify (LLM-Wiki), Refinement Model=DeepRefine-8B2026.05 | 6.69 | |
| HippoRAG2 + DeepRefine-4BConstructor=Graphify (LLM-Wiki), Refinement Model=DeepRefine-4B2026.05 | 6.6 | |
| HippoRAG + DeepRefine-8BConstructor=Graphify (LLM-Wiki), Refinement Model=DeepRefine-8B2026.05 | 6.42 | |
| HippoRAG2Constructor=Graphify (LLM-Wiki)2026.05 | 6.19 | |
| HippoRAG + DeepRefine-4BConstructor=Graphify (LLM-Wiki), Refinement Model=DeepRefine-4B2026.05 | 6.02 | |
| HippoRAGConstructor=Graphify (LLM-Wiki)2026.05 | 5.57 |