Momenta R7 reinforced learning world model mass production launched, Cao Xudong: World model and reinforced learning constitute the two core pillars of physical AI.

On the morning of April 26th, during the recent Beijing International Auto Show, Momenta held a themed event called "Momenta R7, the prologue of Physical AI". Four partners attended the event together, officially announcing the launch of Momenta R7's reinforcement learning world model for mass production. Momenta R7's reinforcement learning world model understands the laws of motion and interactive logic in the physical world, rather than relying on scene memory and rule matching. Momenta announced that it has successfully delivered over 70 mass production models, with a total of over 200 fixed-point models, and mass production has been implemented in over ten countries and regions.