Safe Reinforcement Learning on MetaDrive Medium (held-out seeds 10000–10019)
41Categorical Success RatePPOLag
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| PPOLag2026.06 | 41 | 51 | — | — | |
| VLMPPOLag+Conf2026.06 | 26 | 35 | 26 | — |
| Method | Links | ||||
|---|---|---|---|---|---|
| PPOLag2026.06 | 41 | 51 | — | — | |
| VLMPPOLag+Conf2026.06 | 26 | 35 | 26 | — |