Question Answering on HolisQA Medicine
36.45F1 ScoreHugRAG
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| HugRAGBackbone LLM=gpt-5-nano2026.02 | 36.45 | 69.91 | 60.65 | |
| LeanRAGBackbone LLM=gpt-5-nano2026.02 | 34.25 | 60.43 | 56.6 | |
| CausalRAGBackbone LLM=gpt-5-nano2026.02 | 31.12 | 58.9 | 58.77 | |
| StandardRAGBackbone LLM=gpt-5-nano2026.02 | 26.87 | 61.08 | 56.24 | |
| HippoRAG2Backbone LLM=gpt-5-nano2026.02 | 21.12 | 57.5 | 51.08 | |
| GraphRAG LocalBackbone LLM=gpt-5-nano2026.02 | 19.03 | 56.07 | 49.52 | |
| BM25Backbone LLM=gpt-5-nano2026.02 | 17.72 | 52.04 | 50.64 | |
| GraphRAG GlobalBackbone LLM=gpt-5-nano2026.02 | 17.13 | 54.56 | 48.19 | |
| NaiveGenerationBackbone LLM=gpt-5-nano2026.02 | 12.63 | 0 | 44.7 | |
| LightRAGBackbone LLM=gpt-5-nano2026.02 | 12.16 | 52.38 | 44.15 |