Question Answering on HolisQA Psychology
44.42F1 ScoreHugRAG
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| HugRAGBackbone LLM=gpt-5-nano2026.02 | 44.42 | 60.87 | 63.53 | |
| LeanRAGBackbone LLM=gpt-5-nano2026.02 | 42.85 | 57.46 | 58.65 | |
| CausalRAGBackbone LLM=gpt-5-nano2026.02 | 42.8 | 55.6 | 61.9 | |
| StandardRAGBackbone LLM=gpt-5-nano2026.02 | 37.19 | 52.21 | 59.85 | |
| LightRAGBackbone LLM=gpt-5-nano2026.02 | 33.63 | 45.54 | 56.42 | |
| GraphRAG LocalBackbone LLM=gpt-5-nano2026.02 | 32.31 | 35.22 | 55.02 | |
| GraphRAG GlobalBackbone LLM=gpt-5-nano2026.02 | 31.09 | 34.26 | 54.62 | |
| BM25Backbone LLM=gpt-5-nano2026.02 | 30.46 | 33.44 | 56.63 | |
| NaiveGenerationBackbone LLM=gpt-5-nano2026.02 | 22.91 | 0 | 50 | |
| HippoRAG2Backbone LLM=gpt-5-nano2026.02 | 20.1 | 34.13 | 46.77 |