Get Zlio on your home screen
One tap, full screen, like an app.
A Hugging Face paper proposes factor-aware predictive learning that makes multimodal LLM representations actionable for video diffusion transformers, enabling sparse control of character, scene, action and camera in cinematic video generation.