Dynin-Robotics: Omnimodal Unified Diffusion Vision-Language-Action Model
Dynin-Robotics unifies action, goal, and dynamics prediction in one omnimodal masked-diffusion VLA model, reaching 78.4% success on Franka Research 3 manipulation tasks.
1 stories, newest first · one card per story
Dynin-Robotics unifies action, goal, and dynamics prediction in one omnimodal masked-diffusion VLA model, reaching 78.4% success on Franka Research 3 manipulation tasks.
Today 1.6K readers · 2.5K views · 779 reads · all time 189K views