Maze Navigation on 32x32 Small Mazes (test)
39.6Mean Completion TimeSpiral RL
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Spiral RLConvergence Mode=Q-learning adaptive threshold, Memory Strategy=Full storage2025.12 | 39.6 | 100 | |
| Sentinel RLConvergence Mode=Q-learning adaptive threshold, Memory Strategy=Sampling2025.12 | 40.4 | 100 | |
| Spiral ConvConvergence Mode=Fixed 40% threshold, Memory Strategy=Full storage2025.12 | 45.8 | 100 | |
| Sentinel ConvConvergence Mode=Fixed 40% threshold, Memory Strategy=Sampling2025.12 | 45.8 | 100 | |
| SpiralConvergence Mode=No Convergence, Memory Strategy=Full storage2025.12 | 60 | 100 | |
| SentinelConvergence Mode=No Convergence, Memory Strategy=Sampling2025.12 | 60 | 100 |