Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
This paper investigates whether a shared representational geometry exists across human brains, analogous to universal latent spaces in ANNs. The authors train subject-specific fMRI encoders using a self-supervised approach on the Natural Scenes Dataset, leveraging repeated stimulus presentations. They demonstrate that these independently learned spaces can be aligned across subjects using unsupervised orthogonal rotations, suggesting a shared neural geometry in the human visual cortex.
Human brains might share a universal representational geometry, allowing us to translate neural activity across individuals without paired data or intermediate models.
The Strong Platonic Representation Hypothesis suggests that representational convergence in artificial neural networks can be harnessed constructively: embeddings can be translated across models through a universal latent space without paired data. We ask whether an analogous geometry can be recovered across human brains. Using fMRI data from the Natural Scenes Dataset, we propose a self-supervised encoder that learns subject-specific embeddings from brain data alone by exploiting repeated stimulus presentations. We show that these independently learned spaces can be translated across subjects using unsupervised orthogonal rotations, without paired cross-subject samples or intermediate model representations. Synchronizing pairwise rotations into a single shared latent space further improves cross-subject retrieval, indicating that subject-specific spaces are mutually compatible with a common coordinate system. These results provide evidence for a shared neural geometry in the human visual cortex: subject-specific fMRI representations are approximately isometric across individuals and can be translated through purely geometric transformations.