Crisis response generation on Earthquake and Typhoon cross-crisis scenarios
1Professionalism ScoreFusion w/ Eval & Weight Instruct
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| Fusion w/ Eval & Weight InstructBase Model=Llama-3.1-8B-Instruct2025.09 | 1 | 0.99 | 0.78 | 0.95 | 0.96 | |
| Fusion w/ EvalBase Model=Llama-3.1-8B-Instruct2025.09 | 0.98 | 0.98 | 0.78 | 0.94 | 0.92 | |
| Prompt and SelectBase Model=Llama-3.1-8B-Instruct2025.09 | 0.97 | 0.98 | 0.77 | 0.93 | 0.91 | |
| Fusion w/o EvalBase Model=Llama-3.1-8B-Instruct2025.09 | 0.96 | 0.97 | 0.78 | 0.93 | 0.88 | |
| Fusion w/ Eval & InstructBase Model=Llama-3.1-8B-Instruct2025.09 | 0.96 | 0.97 | 0.78 | 0.93 | 0.89 | |
| RAGBase Model=Llama-3.1-8B-Instruct2025.09 | 0.94 | 0.97 | 0.77 | 0.92 | 0.88 | |
| Instructional PromptBase Model=Llama-3.1-8B-Instruct2025.09 | 0.93 | 0.94 | 0.79 | 0.91 | 0.84 | |
| RAG-PEBase Model=Llama-3.1-8B-Instruct2025.09 | 0.76 | 0.72 | 0.77 | 0.75 | 0.73 |