// OVERVIEW
Developer of Stable Video Diffusion (SVD), a video generation model repurposed for robot forward dynamics modeling.
// RECENT MENTIONS
// SIGNALS
2 SIGNALS
01
mention·arXiv Physical AI·JUNE 11, 2026
“each view is encoded into H×W patch tokens using the pretrained Stable Diffusion 3 VAE encoder”
Source→02
mention·arXiv Physical AI·MARCH 27, 2026
“We adopt the open-sourced Stable Video Diffusion (SVD) (Blattmann et al., 2023) as the general forward dynamics model.”
Source→AI-extracted from podcast / newsletter / paper summaries. May contain errors.