Scientists build a new type of neural network to assist AI in perceiving the world.

Recently, the collaborative research between the Brain Map and Brain-like Intelligence Laboratory of the Institute of Automation, Chinese Academy of Sciences, led by Yu Shan, and the team of Bi Yanchao from the School of Psychology and Cognitive Science at Peking University has made a significant breakthrough. The research proposes a new neural network framework, CATSNet, which achieves human-like concept formation, understanding, and communication. The related findings have been published online in the international academic journal "Nature Computational Science".