Few-shot task expansion on ALFWorld 1-shot (unseen domains)
71.28Success Rate (SR)TMoW
Evaluation results
| Method | Links | ||
|---|---|---|---|
| TMoWshot=12026.01 | 71.28 | 34.23 | |
| SayCanPayshot=12026.01 | 46.08 | 39 | |
| LLM+FTshot=12026.01 | 43.12 | 40.18 | |
| FLAREshot=12026.01 | 12.28 | 43.67 | |
| LLM-Plannershot=12026.01 | 8.91 | 47.43 |