Platonic Representations in the Human Brain: Unsupervised Recovery of Universal Geometry
This work provides evidence for a shared neural geometry in the human visual cortex, enabling cross-subject translation of brain representations, which is a novel finding for neuroscience and brain-computer interfaces.
The paper demonstrates that subject-specific fMRI embeddings from the human visual cortex can be translated across individuals using unsupervised orthogonal rotations, without paired data or intermediate models, supporting the existence of a universal neural geometry.
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.