Safety-MuJoCo Benchmark
收藏arXiv2025-09-30 收录
下载链接:
https://github.com/SafeRL-Lab/Safety-MuJoCo.git
下载链接
链接失效反馈官方服务:
资源简介:
该数据集是一个旨在评估安全强化学习算法性能的基准,它重点关注在奖励和安全指标之间的平衡。该基准不仅包含了与速度相关的约束,还考虑了机器人健康因素,如跌倒和关节限制。任务领域为安全强化学习。
This dataset is a benchmark designed to evaluate the performance of safe reinforcement learning algorithms, with a primary focus on balancing reward and safety metrics. This benchmark not only includes velocity-related constraints but also takes into account robot health-related factors such as falling and joint limits. The task domain falls under safe reinforcement learning.
提供机构:
SafeRL-Lab



