Summarization on GovReport (ROUGE-L)
19.77ROUGE-LHyDE-Rerank
Evaluation results
| Method | Links | |
|---|---|---|
| HyDE-RerankModel=Mistral-7B-Instruct-v0.32026.01 | 19.77 | |
| HyDE-RerankModel=Qwen2-7B-Instruct2026.01 | 19.77 | |
| RAG-RerankModel=Qwen2-7B-Instruct2026.01 | 19.74 | |
| RAG-RerankModel=Mistral-7B-Instruct-v0.32026.01 | 19.7 | |
| S3-HybridModel=Qwen2-7B-Instruct2026.01 | 19.55 | |
| BM25Model=Qwen2-7B-Instruct2026.01 | 19.55 | |
| S3-HybridModel=Mistral-7B-Instruct-v0.32026.01 | 19.31 | |
| BM25Model=Mistral-7B-Instruct-v0.32026.01 | 19.31 | |
| RAG-RerankModel=Llama-3.1-8B-Instruct2026.01 | 19.3 | |
| S3-HybridModel=Llama-3.1-8B-Instruct2026.01 | 19.12 | |
| BM25Model=Llama-3.1-8B-Instruct2026.01 | 19.12 | |
| HyDE-RerankModel=Llama-3.1-8B-Instruct2026.01 | 19.06 | |
| FullKVModel=Qwen2-7B-Instruct2026.01 | 17.74 | |
| FullKVModel=Llama-3.1-8B-Instruct2026.01 | 17.44 | |
| S3-PureModel=Llama-3.1-8B-Instruct2026.01 | 16.93 | |
| RAG-DenseModel=Llama-3.1-8B-Instruct2026.01 | 16.92 | |
| S3-PureModel=Qwen2-7B-Instruct2026.01 | 16.86 | |
| RAG-DenseModel=Qwen2-7B-Instruct2026.01 | 16.86 | |
| S3-PureModel=Mistral-7B-Instruct-v0.32026.01 | 16.44 | |
| RAG-DenseModel=Mistral-7B-Instruct-v0.32026.01 | 16.44 | |
| FullKVModel=Mistral-7B-Instruct-v0.32026.01 | 16.39 |