Scientists from the University of Osnabrück contribute to research on revisiting human vision, as published in Nature Machine Intelligence.

In a groundbreaking study published in the journal Nature Machine Intelligence, researchers from the University of Osnabrück, Montreal, and Minnesota have proposed using artificial intelligence (AI) language models to better understand human vision.

The study, titled "High-level visual representations in the human brain are aligned with large language models," has been making waves in the scientific community. The DOI for the paper is 10.1038/s42256-025-01072-0, and it can be accessed at this link: https://www.nature.com/articles/s42256-025-01072-0.

Prof. Dr. Tim C. Kietzmann from the University of Osnabrück is a co-first author of the study. He and his team trained artificial neural networks to predict language model representations from images. The results suggest that these AI models, which process visual information in a linguistically decodable manner, better represent brain activity in participants than many current leading AI models.

The correspondence between representations in AI language models and activation patterns in the brain is significant for understanding complex semantic processing in the brain. The findings point to possible paths for improving AI systems in the future, with potential applications in medical fields.

One such application could be the development of visual prosthetics for people with severe visual impairments. The new technology could help these individuals interpret visual information more effectively, enhancing their quality of life.

Moreover, this "mind-reading" capability hints at potential improvements for brain-computer interfaces. By understanding how the brain processes visual information in a linguistic context, researchers could develop interfaces that are more intuitive and user-friendly.

Interestingly, the study also suggests that the visual system of the human brain might be trying to find a common language, a lingua franca, across different senses and languages. This would greatly simplify communication between brain regions, facilitating more efficient information processing.

Prof. Dr. Adrien Doerig, who was involved in the study, is now researching at FU Berlin. The first author of the paper, Professor Adrien Doerig from Freie Universität Berlin, can be contacted at [email protected] for press inquiries.

This approach is a novel worldwide, marking a significant step forward in the field of AI and neuroscience. As we continue to explore and understand the intricate workings of the human brain, the potential applications of this research are vast and exciting.