Attributed Question Answering on ALCE
41.023STRREC ScoreVGraphRAG
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| VGraphRAGprotocol=RAG2025.03 | 41.023 | 15.401 | 71.835 | |
| VGraphRAG2025.03 | 41.023 | 15.401 | 71.835 | |
| VGraphRAG-CCprotocol=RAG2025.03 | 37.82 | 13.608 | 68.46 | |
| GraphRAG-CCprotocol=RAG2025.03 | 35.366 | 11.92 | 64.662 | |
| RAPTORprotocol=RAG2025.03 | 35.255 | 11.076 | 65.401 | |
| RAPTORLLM=Llama-3-8B, Embedding model=BGE-M3, top-k=4, chunk size=1,200-token, Clustering method=K-means2025.03 | 35.255 | 11.076 | 65.401 | |
| RAPTOR2025.03 | 35.255 | 11.076 | 65.401 | |
| VanillaRAGprotocol=RAG2025.03 | 34.283 | 11.181 | 63.608 | |
| VanillaRAGLLM=Llama-3-8B, Embedding model=BGE-M3, top-k=4, chunk size=1,200-token2025.03 | 34.283 | 11.181 | 63.608 | |
| FastGraphRAGLLM=Llama-3-8B, Embedding model=BGE-M3, top-k=4, chunk size=1,200-token2025.03 | 30.19 | 8.544 | 56.962 | |
| LGraphRAGprotocol=RAG2025.03 | 28.448 | 8.544 | 54.747 | |
| LGraphRAGLLM=Llama-3-8B, Embedding model=BGE-M3, top-k=4, chunk size=1,200-token2025.03 | 28.448 | 8.544 | 54.747 | |
| LGraphRAG2025.03 | 28.448 | 8.544 | 54.747 | |
| KGPLLM=Llama-3-8B, Embedding model=BGE-M3, top-k=4, chunk size=1,200-token2025.03 | 27.692 | 8.755 | 51.899 | |
| GraphRAG-ERprotocol=RAG2025.03 | 26.774 | 7.5949 | 52.743 | |
| HLightRAGLLM=Llama-3-8B, Embedding model=BGE-M3, top-k=4, chunk size=1,200-token2025.03 | 25.578 | 6.54 | 50.422 | |
| HippoRAGLLM=Llama-3-8B, Embedding model=BGE-M3, top-k=4, chunk size=1,200-token2025.03 | 23.357 | 6.962 | 43.671 | |
| LLightRAGLLM=Llama-3-8B, Embedding model=BGE-M3, top-k=4, chunk size=1,200-token2025.03 | 21.937 | 5.591 | 43.776 | |
| DALKLLM=Llama-3-8B, Embedding model=BGE-M3, top-k=4, chunk size=1,200-token2025.03 | 21.408 | 4.114 | 44.937 | |
| GLightRAGLLM=Llama-3-8B, Embedding model=BGE-M3, top-k=4, chunk size=1,200-token2025.03 | 17.859 | 3.587 | 37.131 | |
| ZeroShotprotocol=Zero-shot2025.03 | 15.454 | 3.692 | 30.696 | |
| ZeroShotLLM=Llama-3-8B2025.03 | 15.454 | 3.692 | 30.696 | |
| ToGLLM=Llama-3-8B, Embedding model=BGE-M3, top-k=4, chunk size=1,200-token2025.03 | 13.975 | 3.059 | 29.114 | |
| G-retrieverLLM=Llama-3-8B, Embedding model=BGE-M3, top-k=4, chunk size=1,200-token2025.03 | 9.754 | 2.215 | 19.726 |