Role-playing Attractiveness on CharacterEval
3.618HLBaichuan2-7B+ICL
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| Baichuan2-7B+ICLBackbone=Baichuan2-7B, Evaluation Mode=ICL2025.03 | 3.618 | 2.71 | 2.095 | 2.964 | 2.847 | |
| Baichuan2-7B+SFT+ICLBackbone=Baichuan2-7B, Evaluation Mode=SFT+ICL2025.03 | 3.521 | 2.841 | 1.948 | 3.106 | 2.854 | |
| Baichuan2-7B+PCLBackbone=Baichuan2-7B, Evaluation Mode=PCL2025.03 | 3.502 | 3.099 | 2.019 | 3.171 | 2.948 | |
| GPT-4+PCL*Backbone=GPT-4, Evaluation Mode=PCL* (Chain of Persona Prompting)2025.03 | 3.267 | 3.245 | 2.212 | 3.243 | 2.992 | |
| GPT-4+ICLBackbone=GPT-4, Evaluation Mode=ICL2025.03 | 3.143 | 3.184 | 2.153 | 3.01 | 2.873 | |
| Qwen-7B+ICLBackbone=Qwen-7B, Evaluation Mode=ICL2025.03 | 3.036 | 2.791 | 2.052 | 2.838 | 2.679 | |
| Qwen-7B+PCLBackbone=Qwen-7B, Evaluation Mode=PCL2025.03 | 2.756 | 3.186 | 1.857 | 3.068 | 2.75 | |
| Qwen-7B+SFT+ICLBackbone=Qwen-7B, Evaluation Mode=SFT+ICL2025.03 | 2.744 | 2.935 | 1.869 | 2.918 | 2.616 | |
| GPT-3.5+PCL*Backbone=GPT-3.5, Evaluation Mode=PCL* (Chain of Persona Prompting)2025.03 | 2.461 | 2.84 | 1.671 | 2.768 | 2.439 | |
| CharacterGLM-6B+ICLBackbone=CharacterGLM-6B, Evaluation Mode=ICL2025.03 | 2.318 | 1.652 | 1.378 | 2.129 | 1.869 | |
| GPT-3.5+ICLBackbone=GPT-3.5, Evaluation Mode=ICL2025.03 | 2.304 | 2.607 | 1.666 | 2.457 | 2.256 | |
| CharacterGLM-6B+PCLBackbone=CharacterGLM-6B, Evaluation Mode=PCL2025.03 | 2.246 | 1.805 | 1.511 | 2.458 | 2.004 | |
| CharacterGLM-6B+SFT+ICLBackbone=CharacterGLM-6B, Evaluation Mode=SFT+ICL2025.03 | 2.233 | 1.714 | 1.365 | 2.286 | 1.9 |