Safe Reinforcement Learning on LTC controller Full oracle access
0ΣδCPO-Style
Evaluation results
| Method | Links | ||
|---|---|---|---|
| CPO-Style2026.03 | 0 | 37 | |
| Safety Shield2026.03 | 0 | 78 | |
| Lipschitz Ballsigma=matched2026.03 | 0 | 0 | |
| Lipschitz Ballsigma=sigma*2026.03 | 0 | 500 | |
| Lyapunov-Style2026.03 | 52 | 37 |