NCCVMay 19

Platonic Representations in the Human Brain: Unsupervised Recovery of Universal Geometry

arXiv:2605.2049658.5
Predicted impact top 14% in NC · last 90 daysOriginality Incremental advance
AI Analysis

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.

Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes