Courtois Project on Neuronal Modelling
The Courtois project on Neural Modelling (cneuromod) aims at training artificial neural networks to mimic extensive experimental data on individual human brain activity and behaviour.
- Reposted by Courtois Project on Neuronal ModellingWe’d love your feedback on BERG (github.com/gifale95/BERG): pretrained encoding models + a Python toolkit for generating in silico neural responses for in silico experimentation. Your input will make BERG more useful and reliable! forms.gle/pybrqcaqdso2... #NeuroAI #CompNeuro #neuroscience #AI
- CNeuroMod is now on substack, and our first post highlights a new study showing that a tiny 2M-parameter audio model can be meaningfully fine-tuned on an individual brain with benefits for downstream AI tasks. open.substack.com/pub/cneuromo...
- Our very own Isil Bilgin will share some updates on her work with Leila Wehbe and @lune-bellec.bsky.social to create brain-aligned language models by fine-tuning LLAMA-7B to produce brain-like embeddings.
- On November 26th noon ET, @complexbrains.bsky.social will present her talk titled " Aligning Brains and Language Models: Toward a NeuroAI Framework for Brain-Grounded Language Understanding". The project combined @cneuromod.ca dataset 🧠🧑🤝🧑 and AI 🤖.
- Brain–AI alignment can reveal exciting similarities in representations. Our Super Mario Bros. experiment demonstrates an important caveat: alignment can be very brittle. 1/🧵
- We showed that the raw memory content of the NES console aligned with brain activity about as well as a convolutional neural network trained with RL. And the brain is obviously not a NES. 2/
- Reposted by Courtois Project on Neuronal Modelling[Not loaded yet]
- In 2019, the CNeuroMod team and 6 participants began a massive data collection journey: twice-weekly MRI scans for most of 5 years. Data collection is now complete! 1/🧵
- The result? Nearly 1000 hours of task fMRI from 6 subjects (mostly 5). We're releasing experiments slowly, focusing on quality checks and well-documented derivatives for reuse. 5/6 participants made their data fully open (CC0), thanks to @conp-pcno.bsky.social
- New CNeuroMod-THINGS open-access fMRI dataset: 4 participants · ~4 000 images (720 categories) each shown 3× (12k trials per subject)· individual functional localizers & NSD-inspired QC . Preprint: arxiv.org/abs/2507.09024 Congrats Marie St-Laurent and @martinhebart.bsky.social !!
- Pair it with ∼75 h of video-watching fMRI per subject used in the 2025 @algonautsproject.bsky.social competition for a one-two controlled-naturalistic punch in modelling individual human vision. Huge thanks: THINGS Initiative, Courtois Foundation, our dedicated participants, and the CNeuroMod crew.
