Reinforcement Learning on MiniGrid FourRooms
27.2Success Rate (SR)No Shaping
Evaluation results
| Method | Links | |
|---|---|---|
| No ShapingTraining Episodes=3,0002026.05 | 27.2 | |
| LLM IterativeTraining Episodes=3,0002026.05 | 23.3 | |
| LLM One-ShotTraining Episodes=3,0002026.05 | 19.9 | |
| Hand-CraftedTraining Episodes=3,0002026.05 | 17.7 | |
| RNDTraining Episodes=3,0002026.05 | 6.1 |