用嘴指挥视频里的角色，只改0.2%参数
现在的视频生成模型已经能听懂人话：你说“让角色向左走、镜头拉近”，它就能照做，但只能整段视频一起动，没法精确到“第3秒到第5秒做这个动作”。这篇把这种粗放的语言控制变成精确的时间控制——把每个指令拆成“角色动作+镜头运动”两部分，再分别绑定到对应的视频时间段，同时加了一个“注意力路由”机制，防止前一个指令的余波污染后一个动作。最反直觉的是，它没给模型加任何新模块，只是把模型在训练时已经学会的语义理解重新利用起来，只训练了0.199%的参数、用8000条游戏视频就实现了。这说明一个趋势：视频模型越强大，语言就越自然地成为控制它的遥控器，而把这种能力变成可用的交互工具，可能不需要从头训练，只需要轻量地“解锁”。它不是你明天能用上的产品，但它是“视频生成变成可玩世界”这条路上一个很轻的脚印。
📄 原文摘要(英文)
We present H3-World, an efficient framework that turns the 33B MiniMax-H3 video generator into an interactive world model. Our key finding is that, as large video generators become more capable, language is emerging as a natural interface for control. MiniMax-H3, for example, already supports zero-shot control of character behavior and camera motion through natural-language instructions. Building on this, H3-World turns this coarse language interface into precise, temporally grounded world control, without introducing dedicated action modules. Specifically, we represent each action as a structured combination of character and camera instructions, and align them with the corresponding temporal video latents. To make the control temporally precise, we further introduce temporal attention routing, which restricts each instruction to its intended time interval and reduces control leakage across actions. Importantly, H3-World directly reuses the semantic representations learned during large-scale video pretraining and requires only lightweight adaptation. With only 8,000 gameplay samples, 10,000 LoRA optimization steps, and 0.199% trainable parameters, H3-World achieves effective character and camera control while preserving strong generation quality. It also generalizes to unseen scenarios. These results show that the control capabilities emerging in large video generators can be efficiently transformed into interactive world control.