Task-Oriented Dialogue on SGD 1.0 (test)
81.29Inform RateDarwinTOD
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| DarwinTODZero-shot=true, Backbone=Qwen3-8B2026.01 | 81.29 | 74.06 | 19.03 | 96.71 | |
| DarwinTODZero-shot=true, Backbone=Qwen2.5-7B2026.01 | 80.71 | 72.16 | 18.86 | 95.3 | |
| DarwinTODZero-shot=true, Backbone=Llama3-8B2026.01 | 80.12 | 71.25 | 19.11 | 94.8 | |
| AgentTODZero-shot=true, Backbone=Llama3-8B2026.01 | 76.7 | 65.82 | 21.66 | 92.92 | |
| ZS-TODZero-shot=true2026.01 | 71.68 | 61.63 | 18.51 | 85.16 | |
| SimpleTODZero-shot=true2026.01 | 68.1 | 60.48 | 20.4 | 84.69 |