Navigation on Two Dimensional Navigation environment (10 seeds)
967.99Final Reward (Avg)Reversibility Aware Control
Evaluation results
| Method | Links | ||
|---|---|---|---|
| Reversibility Aware ControlUsable?=No: Constraints2025.10 | 967.99 | 4 | |
| Baseline PPOUsable?=No: Constraints2025.10 | 959.44 | 4.6 | |
| PPO with SAVMPC shieldUsable?=Yes2025.10 | 842.74 | 0 | |
| PPO with Safety Oracle shieldUsable?=No: Oracle2025.10 | 782.23 | 0 | |
| Leave No TraceUsable?=No: Constraints2025.10 | 375.94 | 6.1 |