Simple Question Answering on PopQA (EM, F1)
56.3F1GTE-Qwen2-7B-Instruct
Evaluation results
| Method | Links | ||
|---|---|---|---|
| GTE-Qwen2-7B-InstructKnowledge Representation=Embedding-based2026.06 | 56.3 | — | |
| GTRKnowledge Representation=Unstructured2026.06 | 56.2 | — | |
| RAPTORKnowledge Representation=Structured2026.06 | 56.2 | — | |
| HippoRAG 2Knowledge Representation=Structured2026.06 | 56.2 | — | |
| SemFlowRAGKnowledge Representation=Structured2026.06 | 56.2 | — | |
| HippoRAGKnowledge Representation=Structured2026.06 | 55.9 | — | |
| GritLM-7BKnowledge Representation=Embedding-based2026.06 | 55.8 | — | |
| NV-Embed-v2 (7B)Knowledge Representation=Embedding-based2026.06 | 55.7 | — | |
| HippoRAG 2 (reproduced)Knowledge Representation=Structured, Reproduction Status=reproduced2026.06 | 55.7 | — | |
| CRAGCategory=Iterative & Agentic RAG2026.04 | 54.5 | 45.5 | |
| ContrieverKnowledge Representation=Unstructured2026.06 | 53.1 | — | |
| Self-Correcting RAGCategory=Self-Correcting RAG2026.04 | 51.9 | 43.2 | |
| FilcoCategory=Advanced Selection & Reranking2026.04 | 51.2 | 42.1 | |
| RAG + MMRCategory=Advanced Selection & Reranking2026.04 | 50.8 | 41.5 | |
| RECOMPCategory=Advanced Selection & Reranking2026.04 | 50.5 | 41.8 | |
| RRRCategory=Standard RAG Baselines2026.04 | 50.1 | 40.2 | |
| BM25Knowledge Representation=Unstructured2026.06 | 49.9 | — | |
| LongLLMLinguaCategory=Advanced Selection & Reranking2026.04 | 49.8 | 40.5 | |
| HyDECategory=Standard RAG Baselines2026.04 | 49.5 | 39.8 | |
| IRCoTCategory=Iterative & Agentic RAG2026.04 | 49.2 | 40.8 | |
| NaiveCategory=Standard RAG Baselines2026.04 | 48.4 | 37.6 | |
| GraphRAGKnowledge Representation=Structured2026.06 | 48.1 | — | |
| Self-RAGCategory=Iterative & Agentic RAG2026.04 | 46.8 | 38.5 | |
| DRAGCategory=Iterative & Agentic RAG2026.04 | 46.5 | 38.6 | |
| Llama-3.3-70B-Instruct w/o RetrievalKnowledge Representation=Unstructured, Retrieval=w/o Retrieval2026.06 | 32.5 | — | |
| LightRAGKnowledge Representation=Structured2026.06 | 2.4 | — |