Question Answering on LiteraryQA
18.4EMHippoRAG2 + CROSSAUG
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| HippoRAG2 + CROSSAUGFramework=HippoRAG2, Setting=CROSSAUG2026.05 | 18.4 | 33.81 | — | — | — | |
| HippoRAG2Framework=HippoRAG2, Setting=Base2026.05 | 17.59 | 32.72 | — | — | — | |
| GFM-RAG + CROSSAUGFramework=GFM-RAG, Setting=CROSSAUG2026.05 | 14.58 | 28.83 | — | — | — | |
| GFM-RAGFramework=GFM-RAG, Setting=Base2026.05 | 12.59 | 27.49 | — | — | — | |
| LightRAG + CROSSAUGFramework=LightRAG, Setting=CROSSAUG2026.05 | 11.63 | 26.82 | — | — | — | |
| LinearRAGFramework=LinearRAG, Setting=-2026.05 | 11.56 | 26.57 | — | — | — | |
| LightRAGFramework=LightRAG, Setting=Base2026.05 | 11.26 | 26.24 | — | — | — | |
| LitSeg2026.05 | 8.1 | 25.6 | 26.1 | 30.9 | 45 | |
| Margin-SamplingRefinement=Merge2026.05 | 8 | 25.5 | 26.2 | 29.9 | 41.8 | |
| PerplexityRefinement=Merge2026.05 | 7.9 | 25.3 | 25.9 | 30.2 | 42.3 | |
| LitSeg-Lite2026.05 | 7.9 | 25.8 | 26.3 | 31.1 | 44.8 | |
| Recursive Char2026.05 | 7.8 | 25.3 | 25.9 | 29.8 | 41.5 | |
| LumberChunkerRefinement=Merge2026.05 | 7.7 | 25.2 | 25.7 | 30.3 | 43.8 | |
| LumberChunker2026.05 | 7.6 | 25 | 25.5 | 29.9 | 43.5 | |
| Perplexity2026.05 | 7.2 | 23.5 | 24.5 | 28 | 37.3 | |
| Token2026.05 | 6.6 | 24 | 24.8 | 29.6 | 41.8 | |
| Margin-Sampling2026.05 | 6.5 | 20.5 | 21.6 | 24.7 | 30.8 |