BrainExplore: Large-Scale Discovery of Interpretable Visual Representations in the Human Brain

TL;DR

This paper introduces BrainExplore, a large-scale automated framework that uncovers and explains interpretable visual representations in the human brain using fMRI data, revealing thousands of distinct visual concepts.

Contribution

The paper presents a novel, scalable method combining unsupervised decomposition and natural language explanations to systematically discover and interpret visual brain representations.

Findings

Revealed thousands of interpretable visual patterns in the human cortex.

Discovered fine-grained visual representations previously unreported.

Developed an automated pipeline for validation and explanation of brain activity patterns.

Abstract

Understanding how the human brain represents visual concepts, and in which brain regions these representations are encoded, remains a long-standing challenge. Decades of work have advanced our understanding of visual representations, yet brain signals remain large and complex, and the space of possible visual concepts is vast. As a result, most studies remain small-scale, rely on manual inspection, focus on specific regions and properties, and rarely include systematic validation. We present a large-scale, automated framework for discovering and explaining visual representations across the human cortex. Our method comprises two main stages. First, we discover candidate interpretable patterns in fMRI activity through unsupervised, data-driven decomposition methods. Next, we explain each pattern by identifying the set of natural images that most strongly elicit it and generating a natural-language description of their shared visual meaning. To scale this process, we introduce an automated pipeline that tests multiple candidate explanations, assigns quantitative reliability scores, and selects the most consistent description for each voxel pattern. Our framework reveals thousands of interpretable patterns spanning many distinct visual concepts, including fine-grained representations previously unreported.