Dialogue Response Generation on P4G (test)
89AccuracyMixed-Initiative Dialogue Prompting
Evaluation results
| Method | Links | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| Mixed-Initiative Dialogue PromptingStrategy=Prompt, Backbone=InstructGPT2023.05 | 89 | 3.83 | 3.71 | 3.69 | 89 | 88 | 3.24 | 59 | 55 | — | |
| RAPStrategy=Fine-tuning (FT)2023.05 | 88 | 3.66 | 3.69 | 3.62 | 87 | 88 | 3.16 | — | 48 | 41 | |
| Ground TruthStrategy=GT2023.05 | 83 | 3.58 | 3.56 | 3.52 | 88 | 88 | 3.09 | 56 | — | 45 |