wan 2.1 文档架构图疑问

<img width="1462" height="534" alt="Image" src="https://GitHub.com/user-attachments/assets/d82d90a5-3e09-4acc-b664-b0435c47803c" /> I looked at the WAN 2.1 paper, and in the architecture diagram, the input video is a panda holding a sign that says "WAN 2.1". After inference calculations, the output generated video looks exactly the same as the input video. Under normal circumstances, adding text control conditions for video generation editing requires that the input and output be different. I don't quite understand what the author intends to convey by using the same input and output videos.