ROBOTERA's VPP2 World Action Model Tops Embodied AI Benchmark RoboDojo

ROBOTERA's VPP2 (Video Prediction Policy 2) World Action Model has achieved the top ranking on the RoboDojo benchmark for general-purpose robot manipulation, demonstrating its ability to connect visual understanding, instruction following, and physical action without additional data or agent-based reinforcement self-improvement. This marks ROBOTERA's fourth benchmark championship in embodied intelligence this year, alongside ongoing humanoid-robot deployments in China.