Adversary Agent Dataset
收藏arXiv2025-09-30 收录
下载链接：
https://github.com/thu-rllab/CFCQL
下载链接
链接失效反馈官方服务：
资源简介：
该数据集由Pan等人（2022年）提供，用于在具有连续动作空间的多人粒子环境中测试CFCQL算法。同时，该数据集还包含了对手智能体，以便对训练后的模型性能进行评估，具体任务是进行多代理强化学习性能评估。
This dataset was provided by Pan et al. (2022) for testing the CFCQL algorithm in multi-agent particle environments with continuous action spaces. Moreover, the dataset contains adversarial agents to support the performance evaluation of the trained models, with the specific task focusing on multi-agent reinforcement learning performance assessment.
提供机构：
Pan et al.
