Proactive Dialogue on DuRecDial ID (test)
4.21ProactivityQwen
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| QwenModel Size=14B, Evaluation Protocol=prompt-based2026.05 | 4.21 | 4.81 | 4.71 | 3.76 | |
| QwenModel Size=32B, Evaluation Protocol=prompt-based2026.05 | 4.06 | 4.81 | 4.76 | 4.05 | |
| LLaMAModel Size=8B, Evaluation Protocol=prompt-based2026.05 | 3.36 | 4.4 | 4.35 | 3.44 | |
| CSMModel Size=0.3B, Evaluation Protocol=fine-tuning2026.05 | 2.85 | 4.41 | 4.29 | 2.55 | |
| QwenModel Size=3B, Evaluation Protocol=fine-tuning2026.05 | 2.44 | 4.35 | 4.39 | 2.4 | |
| LLaMAModel Size=3B, Evaluation Protocol=fine-tuning2026.05 | 2.32 | 4.29 | 4.31 | 2.33 | |
| LLaMAModel Size=1B, Evaluation Protocol=fine-tuning2026.05 | 2.23 | 4.26 | 4.16 | 2.29 |