From Activation to Causality: Discovery of Causal Visual Representations in the Human Brain
2026-05-22 • Computer Vision and Pattern Recognition
Computer Vision and Pattern Recognition
AI summaryⓘ
The authors developed BrainCause, a tool that helps figure out which parts of the brain truly represent specific visual concepts, like faces or places. Instead of just looking at brain areas that light up, BrainCause creates special images that remove or change the concept to test if those brain areas still respond. By doing this causal testing with brain activity predictions and real brain scans, the authors show that many previously identified brain areas might not actually represent the concept, just related features. Their method finds both known and new brain regions that genuinely respond to certain visual concepts.
brain representationfunctional localizationfMRIactivation maximizationcausal testingvisual conceptsimage-to-fMRI encoding modelcounterfactual stimulineurosciencegenerative models
Authors
Yuval Golbari, Navve Wasserman, Matias Cosarinsky, Roman Beliy, Aude Oliva, Antonio Torralba, Michal Irani, Tamar Rott Shaham
Abstract
Identifying which brain regions represent a visual concept in the human brain is a central challenge in neuroscience. Existing approaches have localized coarse functional regions (e.g., faces, places) through activation maximization, identifying regions that activate strongly for a target concept relative to other concepts. Yet strong activation alone does not establish that a region represents the concept itself, as responses may instead be driven by correlated visual or semantic cues. We introduce BrainCause, an automated framework that combines generative and brain models to synthesize controlled stimuli and validate neural representations through targeted causal testing. Given a query specifying a concept of interest, our framework constructs targeted stimulus sets comprising concept images, counterfactual edits that remove the target concept while preserving other image content, and images with candidate correlated distractors. It then uses an image-to-fMRI encoding model to predict brain responses and searches for representations that respond specifically to the target concept over correlated alternatives. BrainCause returns validated candidate representations and proposes follow-up fMRI experiments to further test or extend its discoveries. Our approach successfully recovers known functional localizations and identifies new candidate representations across dozens of concepts, validated on both predicted and measured fMRI data. Critically, we show that without causal validation, a large fraction of localizations would be false positives, confirming that activation alone is insufficient evidence of representation.