NeuroAI talk by Tim Kietzmann
A synergy between machine learning and neuroscience
AI and neuroscience share an intertwined quest: to understand and to build intelligent systems. This joint endeavour has proven beneficial for both disciplines. For neuroscience, AI provides powerful tools for multivariate pattern decoding, end-to-end modelling of neural information processing, AI-guided experimental design, and a novel conceptual framework that influences how we interpret neural processes. For AI, neuroscience offers fresh perspectives on problems that cannot be solved by ‘scaling up’ model and dataset sizes.
Event details from: NeuroAI talk by Tim Kietzmann
- Date
- 11 December 2025
- Time
- 10:00 – 11:00
- Room
- L3.33
In this talk, Prof. Tim Kietzmann will present two projects from his lab, one in each direction of NeuroAI, to illustrate this synergistic relationship. First, he will show that taking inspiration from infant development can yield more robust AI vision systems. Second, he will demonstrate how large language models (LLMs) can be used for interactive brain reading, i.e. how we can integrate brain scans into the token space of LLMs to obtain natural-language access to neural data. Overall, these two projects showcase the promise of NeuroAI for advancing both of its constituent fields.
About Prof. Tim Kietzmann
- BSc/MSc training in Cognitive Science
- PhD jointly supervised at Osnabrück University and Vanderbilt, US (supervisors König and Tong)
- Postdoc in Cambridge with Niko Kriegeskorte, UK
- Assistant professor for AI - Donders Institute Nijmegen, Netherlands
- Full professor for Machine Learning - Institute of Cognitive Science, Osnabrück University.