Search papers, labs, and topics across Lattice.
H3-World transforms the 33B MiniMax-H3 video generator into an interactive world model by leveraging natural language as a control interface for character behavior and camera motion. The framework employs temporal attention routing to ensure precise action timing without the need for dedicated action modules, achieving effective control with minimal training data and parameters. Key findings indicate that this approach not only maintains high generation quality but also generalizes well to previously unseen scenarios, highlighting the potential of large video generators for interactive applications.
Language is becoming a powerful interface for precise control in interactive video generation, enabling effective manipulation with minimal training.
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.
