Jack Gallant
@gallantlab.org
Cognitive, Systems and Computational Neuroscientist, Professor at UC Berkeley, and lab head. Check out our lab web site http://gallantlab.org For the latest news, publications, brain viewers, code and tutorials, and data.
Voxelwise Encoding Models do some things differently from other GLM-related methods because the roots of VEM are in neurophysiology, not psychology. This 2006 paper describes system identification for neurophysiology. Change "neurons" to "voxels" and it all still applies.
tinyurl.com/wu-etal-2006
tinyurl.com/wu-etal-2006
(PDF) Complete functional characterization of sensory neurons by system identification
PDF | System identification is a growing approach to sensory neurophysiology that facilitates the development of quantitative functional models of... | Find, read and cite all the research you need on...
tinyurl.com
January 29, 2026 at 2:41 AM
Voxelwise Encoding Models do some things differently from other GLM-related methods because the roots of VEM are in neurophysiology, not psychology. This 2006 paper describes system identification for neurophysiology. Change "neurons" to "voxels" and it all still applies.
tinyurl.com/wu-etal-2006
tinyurl.com/wu-etal-2006
Are you on neurotree.org? This site was started @ UCB by my former grad students Stephen David (now Prof at OHSU) and Ben Hayden (now Prof at Baylor), and they've run it ever since (see tinyurl.com/neurotree2026 for history). Please keep your neurotree info updated, it is a great resource!
Neurotree
Neurotree - The Neuroscience Academic Family Tree
neurotree.org
January 22, 2026 at 1:31 AM
Are you on neurotree.org? This site was started @ UCB by my former grad students Stephen David (now Prof at OHSU) and Ben Hayden (now Prof at Baylor), and they've run it ever since (see tinyurl.com/neurotree2026 for history). Please keep your neurotree info updated, it is a great resource!
A recent paper made big waves arguing that fMRI data are hopelessly confounded and uninterpretable. But I'm always suspicious of strong claims that go against a large body of well-supported science. This important post by @alexanderhuth.bsky.social suggests that the paper is fundamentally flawed.
This paper had a pretty shocking headline result (40% of voxels!), so I dug into it, and I think it is wrong. Essentially: they compare two noisy measures and find that about 40% of voxels have different sign between the two. I think this is just noise!
Would love to hear expert views on this paper. It appears to show that the operationalization of brain activity the field has relied on for 3 decades—the BOLD response—is not actually a sensible measure of brain activity.
www.nature.com/articles/s41...
www.nature.com/articles/s41...
January 20, 2026 at 9:17 PM
A recent paper made big waves arguing that fMRI data are hopelessly confounded and uninterpretable. But I'm always suspicious of strong claims that go against a large body of well-supported science. This important post by @alexanderhuth.bsky.social suggests that the paper is fundamentally flawed.
Put this new preprint by T. Zhang on your reading list! People drove a car to navigate in VR. Voxelwise encoding models were fit using 38 feature spaces (28,134 features!). We show the cortical navigation network comprises 11 regions arranged in functional gradients.
www.biorxiv.org/content/10.6...
www.biorxiv.org/content/10.6...
January 6, 2026 at 3:50 PM
Put this new preprint by T. Zhang on your reading list! People drove a car to navigate in VR. Voxelwise encoding models were fit using 38 feature spaces (28,134 features!). We show the cortical navigation network comprises 11 regions arranged in functional gradients.
www.biorxiv.org/content/10.6...
www.biorxiv.org/content/10.6...
Voxelwise Encoding Models (VEMs) are a great framework for modeling fMRI data, but it can be difficult to implement. We've made VEM accessible by providing software, tutorials and reviews that guide its use an implementation. Get it here:
gallantlab.org/blog/2025-12...
#neuroscience, #neuroimaging
gallantlab.org/blog/2025-12...
#neuroscience, #neuroimaging
December 23, 2025 at 8:31 PM
Voxelwise Encoding Models (VEMs) are a great framework for modeling fMRI data, but it can be difficult to implement. We've made VEM accessible by providing software, tutorials and reviews that guide its use an implementation. Get it here:
gallantlab.org/blog/2025-12...
#neuroscience, #neuroimaging
gallantlab.org/blog/2025-12...
#neuroscience, #neuroimaging