From Activation to Causality: Discovery of Causal Visual Representations in the Human Brain
Summary
BrainCause framework uses generative and brain models to identify causal neural representations in the human brain, demonstrating that activation alone is insufficient for confirming concept representation.
View Cached Full Text
Cached at: 06/03/26, 03:38 PM
Paper page - From Activation to Causality: Discovery of Causal Visual Representations in the Human Brain
Source: https://huggingface.co/papers/2605.23895
Abstract
BrainCause framework uses generative and brain models to identify valid neural representations through causal testing, demonstrating that activation alone is insufficient for confirming concept representation.
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) throughactivation 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 andbrain modelsto synthesize controlled stimuli and validate neural representations through targetedcausal 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 animage-to-fMRI encoding modelto 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 knownfunctional localizationsand 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.
View arXiv pageView PDFProject pageAdd to collection
Get this paper in your agent:
hf papers read 2605\.23895
Don’t have the latest CLI?curl \-LsSf https://hf\.co/cli/install\.sh \| bash
Models citing this paper0
No model linking this paper
Cite arxiv.org/abs/2605.23895 in a model README.md to link it from this page.
Datasets citing this paper1
#### BrainCause/Concept_Targeted_Causal_Images Viewer• Updated4 minutes ago • 208k • 1.14k • 2
Spaces citing this paper0
No Space linking this paper
Cite arxiv.org/abs/2605.23895 in a Space README.md to link it from this page.
Collections including this paper2
Similar Articles
@MSFTResearch: Researchers introduce generative causal testing, which translates black box models into clear hypotheses and verifies t…
Microsoft Research and collaborators introduce generative causal testing (GCT), a method that distills black-box brain prediction models into testable explanations and validates them with fMRI experiments, revealing specific brain region responses to language concepts.
Causal Probing for Internal Visual Representations in Multimodal Large Language Models
This paper proposes a causal framework for probing internal visual representations in Multimodal Large Language Models, revealing differences in how entities and abstract concepts are encoded. The study highlights that increasing model depth is crucial for encoding abstract concepts and uncovers a disconnect between perception and reasoning in current MLLMs.
Causal Discovery in the Era of Agents
This paper argues that language model agents should assist causal discovery workflows by providing contextual support and explanations rather than generating causal conclusions, and introduces causal-learn+ platform to demonstrate this principle.
Introducing Activation Atlases
OpenAI introduces Activation Atlases, a technique for visualizing and understanding the internal representations of neural networks, enabling humans to discover spurious correlations and unexpected behaviors such as fooling image classifiers by adding noodles to images.
CausaLab: A Scalable Environment for Interactive Causal Discovery Toward AI Scientists
CausaLab is a scalable environment for evaluating LLM agents on interactive causal discovery, assessing both predictive accuracy and faithful recovery of underlying causal mechanisms. Experiments reveal a gap between prediction and mechanism recovery, highlighting limits in current LLM agents as experimental causal reasoners.