Question Answering on COVID-QA RAGBench (test)
38.23FC MeanUnWeaver
Evaluation results
| Method | Links | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| UnWeaverLLM=GPT-oss-120B, Embedder=Qwen3-Embedding-4B, Query Prompt=Fixed2026.02 | 38.23 | 1.32 | 1 | 994 | 4 | 230 | 2 | 17 | |
| VectorRAGLLM=GPT-oss-120B, Embedder=Qwen3-Embedding-4B, Query Prompt=Fixed2026.02 | 37.9 | 6.02 | — | 922 | — | 359 | 476 | 17 | |
| RAPTORLLM=GPT-oss-120B, Embedder=Qwen3-Embedding-4B, Query Prompt=Fixed2026.02 | 36.43 | 8.04 | 770 | 2 | 389 | 305 | 730 | 17 | |
| GraphRAGLLM=GPT-oss-120B, Embedder=Qwen3-Embedding-4B, Query Prompt=Fixed2026.02 | 30.92 | 3.66 | 25 | 7 | 14 | 282 | 20 | 17 |