用于训练通用智能体的大规模多智能体迷你游戏
Massively Multiagent Minigames for Training Generalist Agents
AI总结：
本文提出Meta MMO基准，通过多个计算高效的迷你游戏扩展Neural MMO，并研究用单一权重集在多个迷你游戏上的泛化能力，以促进多智能体通用智能体的训练与评估。
摘要：
我们提出了Meta MMO，这是一个由多个多智能体迷你游戏组成的集合，用作强化学习基准。Meta MMO构建在Neural MMO之上，后者是一个大规模多智能体环境，曾是两届NeurIPS竞赛的主题。我们的工作通过引入若干计算高效的迷你游戏扩展了Neural MMO。我们通过学习用一组权重来玩多个迷你游戏，探索了Meta MMO中的泛化能力。我们在MIT许可证下发布了该环境、基线和训练代码。我们希望Meta MMO能够推动Neural MMO的进一步进展，并且更广泛地，能够作为一个有用的多智能体泛化基准。
英文摘要：
We present Meta MMO, a collection of many-agent minigames for use as a reinforcement learning benchmark. Meta MMO is built on top of Neural MMO, a massively multiagent environment that has been the subject of two previous NeurIPS competitions. Our work expands Neural MMO with several computationally efficient minigames. We explore generalization across Meta MMO by learning to play several minigames with a single set of weights. We release the environment, baselines, and training code under the MIT license. We hope that Meta MMO will spur additional progress on Neural MMO and, more generally, will serve as a useful benchmark for many-agent generalization.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
