[DSRP Evidence](https://dsrpevidence.org/)

# Evidence for compositionality in fMRI visual representations via Brain Algebra

Ferrante et al., 2025, Communications Biology — Neuroscience

Patterns: [Systems](https://dsrpevidence.org/pattern/systems)

## In short

A whole scene's neural representation can be rebuilt by algebraically combining its parts.

## What they found (results)

Using latent diffusion models applied to fMRI, neural representations of complex visual stimuli could be decomposed into and recombined from component concepts, so combining part-representations predicts the whole-representation in visual cortex.

## What they set out to do (purpose)

To test whether human visual cortical representations are compositional — wholes built from separable, recombinable parts.

## In more detail

If the brain's representation of a whole scene can be reconstructed by algebraically combining the representations of its constituent parts, then the whole is literally composed of its parts. Purpose: To test whether human visual cortical representations are compositional — wholes built from separable, recombinable parts. Finding: Using latent diffusion models applied to fMRI, neural representations of complex visual stimuli could be decomposed into and recombined from component concepts, so combining part-representations predicts the whole-representation in visual cortex. [Added 2026-07-30 as new evidence beyond the original 109. DSRP structural basis: S (Compositional Identity): a whole is the integrated combination of its parts.]

## Abstract

Electrophysiological and neuroimaging studies have revealed how the brain encodes various visual categories and concepts. An open question is how combinations of multiple visual concepts are represented in terms of the component brain patterns: are brain responses to individual concepts composed according to algebraic rules? To explore this, we generated "conceptual perturbations" in neural space by averaging fMRI responses to images with a shared concept (e.g., "winter" or "summer"). After thresholding to ensure specificity, we applied these perturbations to the neural pattern associated with a base image, forming new brain patterns that incorporate the added concept. These modified brain patterns were then decoded into images using a pretrained fMRI-to-image decoding model. Qualitative and quantitative inspection of the resulting images provides insight into how the brain might combine visual concepts. For example, adding a "winter" perturbation to the brain pattern of a man on a skateboard yields a new pattern representing a man on a snowboard in a winter scene-even when the perturbation modifies only a small subset of voxels. Our findings reveal that compositional processes in neural representations may lead to predictable perceptual outcomes, as interpreted by our decoding model. This suggests that the brain's combinatory encoding of concepts may follow a systematic, algebraic-like process-what we term "brain algebra." Although our study is model-driven, it opens avenues for future empirical work into the mechanisms of compositionality in the brain.

These researchers were not testing DSRP. The finding is theirs; the correspondence to DSRP is drawn by this site.

[Source](https://doi.org/10.1038/s42003-025-08706-4)