Excited to share CortexMAE, a family of fMRI foundation models.
We show that our fMRI foundation models scale predictably with data and compute, following clear power-law behavior in one of the most comprehensive scaling studies in neuroimaging.
Our models also achieve SOTA performance in decoding individuals’ dynamic mental states.
NEW RELEASE:
Today we're releasing CortexMAE: a family of fMRI foundation models trained on 2.1K hours of open fMRI data.
We're also releasing Brainmarks: an open benchmark suite for evaluating fMRI foundation models.
Full paper is on arXiv (accepted to ICML 2026)
A thread:
A super long overdue (3+ years?) post on scaling laws.
Compute is expensive. Scaling laws are a way to help us reason about the optimal compute allocation between data and model size before committing to a large run.
The post covers what scaling laws predict, how compute-optimal allocation works, why Kaplan et al. and Chinchilla disagree, and how data limits + fitting details make extrapolation tricky.
https://t.co/HP26eJvjHB
Soccer fans visiting a very large gas station for the first time... carry-on BBQ sandwiches & beaver nuggets? Yes. An actual beaver? ABSOLUTELY NOT!