Neuroscience

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

Pablo Marcos-Manchón, Rishi Jha, Lluís Fuentemilla

Featured June 6, 2026

This analysis was generated by SciGrove. Upload your own PDFs or enter a DOI — and get the same AI breakdown on any paper.

Get started

AI-generated analysis — This is SciGrove's AI interpretation of the paper, not peer-reviewed content. Always refer to the original paper.

Simply

Brains looking at different pictures still organize visual information in surprisingly similar ways, so this paper shows how to rotate each person's brain map to match a universal brain map, even without seeing the same pictures.

In depth

The paper extends the Strong Platonic Representation Hypothesis to human brains, demonstrating that fMRI representations from different individuals can be aligned without shared data. It introduces a self-supervised encoder to learn subject-specific embeddings and then uses unsupervised orthogonal rotations to translate these spaces into a common coordinate system, revealing an approximately isometric shared neural geometry.

Key Takeaways

1
Subject-specific fMRI embeddings can be learned self-supervised from repeated stimulus presentations alone, without external models or cross-subject supervision.
2
Independently learned brain spaces are approximately isometric across individuals, allowing accurate translation using simple unsupervised orthogonal rotations.
3
Synchronizing pairwise rotations into a single shared latent space significantly improves cross-subject retrieval, confirming that individual brain spaces are compatible with a common coordinate system.

Conceptual Flow

HIGH LEVEL

Methodology: Aligning Brain Maps

The method first makes a unique map for each person's brain, then figures out how to twist and turn these maps so they all line up perfectly, like solving a puzzle without a picture.

Each Person's Brain Data

Create Unique Maps

Aligned Brain Maps

Results: Shared Brain Understanding

They found that brain maps from different people can indeed be lined up very well, showing that everyone's brain uses a similar hidden structure to understand the world.

Different Brain Maps

Find Common Structure

Shared Brain Understanding