Qianjue releases Polibrain predictive world model for robots

In 2026, Qianjue Technology unveiled the predictive world model Polibrain at the Physical AI and World Model Conference. The model enables robots to predict and plan by learning physical laws and environmental state changes. Compared to visual prediction, Polibrain focuses more on physical changes and key variables, and has been applied to quadruped, humanoid, wheeled, and dual-arm robot forms. The research team introduced methods such as polynomial representation and effective order to enhance model generalization and explored cross-entity data reuse, providing new directions for the industrial application of embodied intelligence.

Qianjuan Technology, a Chinese company affiliated with Tsinghua University, has released Polibrain, a prediction-based world model designed to enable robots to understand and predict physical changes in their environment. The model is intended for use across various robot forms, including quadruped, humanoid, and wheeled robots. The development is based on research that predates similar work by LeCun's team, and the model incorporates polynomial representation and causal memory to improve generalization and task execution across different robot morphologies.

A Chinese company affiliated with Tsinghua University has released a prediction-based world model called Polibrain, designed to enable robots to predict and plan actions in dynamic environments. The model focuses on state representation and generalization across different robot morphologies, with a theoretical foundation in polynomial representation and causal memory. The research is positioned as a contribution to the broader field of embodied AI and world modeling, with potential applications in various robotic tasks.