SafeVLA: Towards Safety Alignment of Vision-Language-Action Model via Safe Reinforcement Learning

Re-assign community