The reporter learned from the Institute of Automation of the Chinese Academy of Sciences that Li Guoqi, a researcher of the Institute, and Tian Yonghong, a professor of Peking University, cooperated to build and open source the deep pulse neural network learning framework "Jingzhe", which can provide a full stack of deep pulse learning solutions, and will strongly promote the research and application of deep pulse learning. The relevant achievements were recently published online in the international academic journal "Progress in Science". Pulse neural networks are known as the third generation neural networks and have attracted much attention due to their sparse computing, event driven, and ultra-low power consumption characteristics. With the introduction of deep learning methods, the performance of pulse neural networks has been greatly improved, and pulse deep learning has become an emerging research hotspot, which is of great significance for the development of artificial intelligence. Li Guoqi introduced that traditional pulse neural network frameworks focus more on biological interpretability and do not support automatic differentiation, which cannot fully utilize the large-scale parallel computing capabilities of GPUs, and also lack support for neural morphology sensors and computing chips. The "Jingzhe" framework largely solves the above problems, supporting multiple functions such as neural morphology data processing, deep pulse neural network construction, and neural morphology chip deployment. Compared to traditional pulse neural network frameworks, the 'Jingzhe' framework has the characteristics of simplicity, strong scalability, and high performance. By utilizing the 'Jingzhe' framework, researchers can quickly learn and use across domains, and easily construct and train deep pulse neural networks, "said Tian Yonghong. (New News Agency)