NeurIPS 2025 · Conference paper

Bridging Scales: Spectral Theory Reveals How Local Connectivity Rules Sculpt Global Neural Dynamics in Spatially Extended Networks

Yuhan Huang, Keren Gao, Dongping Yang, Sen Song, Guozhang Chen

Spectral theory links microscopic excitation-inhibition connectivity rules to eigenvalue signatures that predict macroscopic cortical waves, oscillations, and chaos.

The scientific question

How do local synaptic connectivity rules sculpt global spatiotemporal brain dynamics?

Cortical activity shifts across arousal states between asynchronous firing, oscillations, bumps, and waves, yet connecting local circuit parameters to macroscopic regimes remained difficult.

Spatial network structure and connection statistics.
Figure 1 from the paper ↗

The core idea

Spatial Connectivity Spectral Theory

We derive an analytical spectral theory for spatially extended excitatory-inhibitory networks, mapping projection ranges, connection strengths, and densities onto bulk disk and outlier eigenvalues.

What we found

Analytical Prediction of Dynamical Regimes

The eigenvalue spectrum quantitatively predicts transitions among synchronous, asynchronous, oscillatory, bump, traveling-wave, and chaotic regimes, offering a principled starting point for inferring effective connectivity changes from macroscopic brain activity across arousal states.

Continue the question

Related reading

Explore our research

Cite this work

Yuhan Huang, Keren Gao, Dongping Yang, Sen Song, Guozhang Chen. Bridging Scales: Spectral Theory Reveals How Local Connectivity Rules Sculpt Global Neural Dynamics in Spatially Extended Networks. NeurIPS 2025. doi:10.52202/085713-2217.