Agora-2 · 项目原始材料
核对原文摘录 · 4 段，支持 6 项资料
原阅读方向 / 一句话认识 / 输出什么
A simulation model predicts how human and agent actions change the world, while a rendering model translates those changes into generated video for each player.输入什么
given the current world state, recent motion history, and player and agent actions, it predicts the changes that the world server applies at the next simulation step.动作是否影响预测
The simulation model is a learned state-transition model: given the current world state, recent motion history, and player and agent actions, it predicts the changes that the world server applies材料说明的限制
Long-horizon visual quality, agreement across views, end-to-end action adherence, and agent-policy performance remain unevaluated in this report.