Source-linked AI summary
Transition to chaos in random networks with cell-type-specific connectivity
Johnatan Aljadeff, Merav Stern, Tatyana O. Sharpee
TL;DR
Conventional firing-rate models assume one connectivity rule, although neuronal types differ in excitability and connectivity. The paper extends mean-field analysis to cell-type-specific networks, finding a silent-to-chaotic transition tied to connectivity structure and block-structured random-matrix spectra.
Problem
Conventional firing-rate models assume one connectivity rule, while experiments highlight neuron-type diversity in excitability and connectivity.
Source: p. 1Method
The paper extends firing-rate models and mean-field analysis to multiple cell types with cell-type-dependent connectivity.
Source: p. 1Results
The transition depends on connectivity structure; locating it yields a result on the spectral-density support of asymmetric matrices with block-structured variances.
Source: p. 1Takeaways & Limitations
Earlier studies suggest computational capacity is optimal near this transition, linking cell-type-specific connectivity to a capacity-relevant regime.
Source: p. 1Applying the results to externally driven, nonautonomous multi-type networks is suggested by analogy, rather than analyzed directly here.
Source: p. 4
Abstract
from arXiv · showhide
In neural circuits, statistical connectivity rules strongly depend on neuronal type. Here we study dynamics of neural networks with cell-type specific connectivity by extending the dynamic mean field method, and find that these networks exhibit a phase transition between silent and chaotic activity. By analyzing the locus of this transition, we derive a new result in random matrix theory: the spectral radius of a random connectivity matrix with block-structured variances. We apply our results to show how a small group of hyper-excitable neurons within the network can significantly increase the network's computational capacity.