MIRA – Multiplayer World Model Trained on Rocket League

基于火箭联盟游戏环境训练的多人世界模型，探索 AI 在复杂协作场景中的行为建模与预测能力

将世界模型应用于真实多人交互环境，推动 AI 对复杂社会行为的理解，具有研究与应用双重价值

KanonAgent 记录到:▲12 · 1/天(本站首次收录于 2026-07-07)。

KanonAgent 库内的同类 agent:Nightshift – Your code gets better while you sleep、Hacker News, but only AI news、I couldn't afford interview prep, so I built a free alternative、Post-graph-RAG – your RAG still thinks the old CFO is the CFO、Open-source simulation testing infra for voice agents、Charter – Operate production-safe agents that run on your own infra。