General Question Answering on Aggregated Benchmarks Average
61.2Average F1SemFlowRAG
Evaluation results
| Method | Links | |
|---|---|---|
| SemFlowRAGKnowledge Representation=Structured2026.06 | 61.2 | |
| HippoRAG 2Knowledge Representation=Structured2026.06 | 59.8 | |
| HippoRAG 2 (reproduced)Knowledge Representation=Structured, Reproduction Status=reproduced2026.06 | 59.6 | |
| NV-Embed-v2 (7B)Knowledge Representation=Embedding-based2026.06 | 57 | |
| GritLM-7BKnowledge Representation=Embedding-based2026.06 | 56.1 | |
| GTE-Qwen2-7B-InstructKnowledge Representation=Embedding-based2026.06 | 54.9 | |
| HippoRAGKnowledge Representation=Structured2026.06 | 53.1 | |
| GTRKnowledge Representation=Unstructured2026.06 | 50.4 | |
| GraphRAGKnowledge Representation=Structured2026.06 | 49.6 | |
| RAPTORKnowledge Representation=Structured2026.06 | 48.8 | |
| BM25Knowledge Representation=Unstructured2026.06 | 47.7 | |
| ContrieverKnowledge Representation=Unstructured2026.06 | 46.9 | |
| Llama-3.3-70B-Instruct w/o RetrievalKnowledge Representation=Unstructured, Retrieval=w/o Retrieval2026.06 | 38.4 | |
| LightRAGKnowledge Representation=Structured2026.06 | 6.6 |