**[AI "Mind-Reading" Makes New Progress: Brain-IT Can Reconstruct Seen Images Based on Brain Activity]** Golden Finance reported on October 10 that the Weizmann Institute of Science in Israel recently announced that its research team has developed an artificial intelligence model called Brain-IT, which can reconstruct image content based on functional magnetic resonance imaging (fMRI) data collected when humans view images. Unlike existing methods that typically require tens of hours of brain scan data, this model can learn to decode the brain activity of new subjects in just about one hour and more accurately restore details such as image content, composition, and color. The research team developed image encoders and decoders and used artificially generated predictive brain activity data to address the issue of insufficient training samples. During the research process, the encoder identified 128 functional regions shared across different populations for image processing, some of which had not been previously discovered. The research team stated that studies related to Brain-IT have been selected for presentation at the International Conference on Learning Representations (ICLR). Currently, the team is exploring the extension of this technology to decode auditory information and investigating its potential application for video decoding in the future.