NeuroAI: cross-pollination between Neuroscience and Artificial Intelligence
Presentation
NeuroAI embodies the cross-pollination between Neuroscience and Artificial Intelligence (AI). On the one hand, the team leverages the staggering recent breakthroughs in AI tools to model brain processing more accurately – for example, visual processing or mental simulations – or to find patterns in vast amounts of fMRI, EEG, and MEG data, and to relate them to stimulus features, perception, cognition, behavior, and well-being. On the other hand, all the recent AI models are black boxes, still limited in generalization, and have an enormous computational cost (energy, chips, data, etc.). Therefore, another objective is to seek inspiration from the brain to design more interpretable, robust, and frugal AIs, for example, by incorporating spikes, more human-like visual representations, or a cognitive architecture, presumably used by the brain and known as the global workspace.
Projects
Neuro-AI guided objective hearing assessment and hearing loss compensation
The Global Latent Workspace: Towards AI models of flexible cognition
Cobots with Conversation, Cognition and Perception
Tuning EEG Fundation Models to Brain Dynamics
Experience-Shaped Geometry of Cognition : The Riemannian Mind Hypothesis
A Predictive Coding Perspective on Brain Dynamics: the case of oscillatory traveling waves
Publications (selection)
- Béthune, L., Vigouroux, D., Du, Y., VanRullen, R., Serre, T., & Boutin, V. (2025). Follow the Energy, Find the Path: Riemannian Metrics from Energy-Based Models. NeuRIPS.
- Ferrante, M., Boccato, T., Toschi, N., & VanRullen, R. (2025). Evidence for compositionality in fMRI visual representations via Brain Algebra. Communications Biology.
- Schwenk, J.C.B., & Alamia, A. (2025). A hierarchical multiscale model of forward and backward alpha-band traveling waves in the visual system. PLOS Computational Biology.
- Rançon, U., Masquelier, T., Cottereau, B.R. (2025). Temporal recurrence as a general mechanism to explain neural responses in the auditory system. Communications Biology.
- Hammouamri, I., Khalfaoui-Hassani, I., Masquelier, T. (2024). Learning Delays in Spiking Neural Networks using Dilated Convolutions with Learnable Spacings. ICLR.
- Devillers, B., Maytié, L., & VanRullen, R. (2024). Semi-supervised multimodal representation learning through a global workspace. IEEE Transactions on Neural Networks and Learning Systems.
- Alamia, A., Gordillo, D., Chkonia, E., Roinishvili, M., Cappe, C., & Herzog, M. H. (2024). Oscillatory traveling waves provide evidence for predictive coding abnormalities in schizophrenia. Biological Psychiatry.
- Boutin, V., Mukherji, R., Agrawal, A., Muzellec, S., Fel, T., Serre, T. and VanRullen, R. (2024). Latent Representation Matters: Human-like Sketches in One-shot Drawing Tasks. NeuRIPS.
- Delorme, A. (2023). EEG is better left alone. Scientific Reports.
- Truong, D., Khalid, M. A., & Delorme, A. (2023). Deep learning applied to EEG data with different montages using spatial attention. IEEE BIBM.
Funding
