Multi-Agent Environment

The dataset used in the paper is a multi-agent environment where agents learn to coordinate their actions to achieve a common goal. The dataset is used to evaluate the proposed method of using a generative model to infer missing observations during decentralized execution.

本文所使用的数据集为多智能体环境（multi-agent environment），在该环境中智能体需学习协调自身行动以达成共同目标。该数据集被用于评估所提出的利用生成式模型（generative model）在分布式执行（decentralized execution）过程中推断缺失观测值的方法。