AI Insight
This study develops a mathematical framework for understanding how neural populations encode sensory information by constructing a geometry where distances between stimuli reflect how reliably they can be distinguished from neural activity. The researchers derive a multi-scale extension of Fisher information metric from first principles that directly relates to mutual information, such that stimulus directions contributing more information are geometrically expanded while poorly encoded directions contract. When applied to visual cortex responses to natural images using diffusion models for practical estimation, the method identifies interpretable stimulus features that most contribute to encoded information.
Why it matters
This framework provides a principled, information-theoretic approach to decode what information neural populations carry about sensory stimuli, with practical applications for analyzing large-scale neural recordings. The method could improve brain-computer interfaces and artificial neural network design by identifying which stimulus features are most robustly encoded by biological neural systems.
Understand the Science
⚠️ Preprint – Noch nicht peer-reviewed
Dieser Artikel wurde noch nicht von unabhängigen Experten begutachtet. Die Ergebnisse sind vorläufig und sollten mit Vorsicht interpretiert werden.
Abstract: Understanding how neural population responses represent sensory information is a central problem in systems neuroscience. One approach is to define a representational geometry on stimulus space in which distances reflect how reliably stimuli can be distinguished from neural activity. However, different constructions of these distances can lead to qualitatively different conclusions about the neural code. Here, we show that a unique Riemannian representational geometry emerges from first principles governing how distances contract as stimulus resolution is lost through coarse-graining. This results in a multi-scale extension of the Fisher information metric, capturing encoding structure from fine stimulus details to coarse global distinctions. The resulting geometry is exactly related to the mutual information encoded by the population: well-encoded stimulus directions (those contributing more to mutual information) are expanded, whereas poorly encoded directions are contracted. The metric tensor can be estimated using diffusion models, making the framework practical for large neural populations and high-dimensional stimuli. Applied to visual cortical responses to natural images, the eigenvectors of the metric tensor identify stimulus variations that contribute most to the encoded stimulus information, yielding interpretable features that are robust to modelling choices. Together, these results provide a principled, information-theoretic framework for characterising neural population codes.
Source: A multi-scale information geometry reveals the structure of mutual information in neural populations