Ep#106: Flex-π: A Multi-Stream World-Action Model with Compute Flexibility

Podcast: RoboPapers · Chris Paxton and Michael Cho

Opublikowano: 2026-09-23

World Action Models are becoming more popular in robotics, as they learn to predict the world jointly with learning how to act on it. However, these world predictions are usually purely based on reconstructing color images from video. This is limiting, because color is far from the most important quality for a robot moving around in the world — more important are qualities like 3D geometry and object semantics. In Flex-π, Ge Yan, Jesse Zhang, and team train a 6 billion parameter world action model to do exactly this, predicting 3D pointmaps and DINO features along with color. This results in a p...