Multi-turn role-play on MedD
95.3Success Rate (SR)EmoLLM
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| EmoLLMBackbone=Qwen3-8B, Method Category=Ours2026.03 | 95.3 | 2.18 | 4.08 | 4.59 | |
| PERMBackbone=Qwen3-8B, Method Category=RL Methods2026.03 | 94 | 2.28 | 4.1 | 4.18 | |
| RLVERBackbone=Qwen3-8B, Method Category=RL Methods2026.03 | 93 | 2.36 | 4.08 | 4.01 | |
| ICL-AIFBackbone=Qwen3-8B, Method Category=Prompting Methods2026.03 | 92.7 | 2.31 | 4.06 | 4.06 | |
| RLFF-ESCBackbone=Qwen3-8B, Method Category=RL Methods2026.03 | 92.4 | 2.44 | 4.02 | 3.92 | |
| ARG-PromptBackbone=Qwen3-8B, Method Category=Prompting Methods2026.03 | 92.3 | 2.28 | 4.02 | 4.01 | |
| PPDPBackbone=Qwen3-8B, Method Category=SFT Methods2026.03 | 91.8 | 2.37 | 4 | 3.89 | |
| ESCoTBackbone=Qwen3-8B, Method Category=SFT Methods2026.03 | 91.2 | — | 4.01 | 3.76 | |
| Best-of-NBackbone=Qwen3-8B, Method Category=Prompting Methods2026.03 | 90.8 | 2.41 | 3.92 | 3.97 | |
| Ask-an-ExpertBackbone=Qwen3-8B, Method Category=Prompting Methods2026.03 | 90.8 | 2.63 | 3.96 | 3.72 | |
| SFTBackbone=Qwen3-8B, Method Category=SFT Methods2026.03 | 90.4 | 2.29 | 3.91 | 3.68 | |
| Qwen3-8BBackbone=Qwen3-8B, Method Category=Base Model2026.03 | 89.2 | 2.92 | 3.95 | 3.57 |