Planning on TravelPlanner (test)
0.271Success RateAutoRefine
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| AutoRefineBackbone=GPT-4-turbo2026.01 | 0.271 | 21.8 | |
| ReActBackbone=GPT-4-turbo2026.01 | 0.104 | 26.1 | |
| ReAct + ReflexionBackbone=GPT-4-turbo2026.01 | 0.091 | 80.2 | |
| ReflexionBackbone=GPT-4-turbo2026.01 | 0.086 | 77.9 |