Stochastic activity in low-rank recurrent neural networks
Open Access
- 18 August 2025
- journal article
- research article
- Published by Public Library of Science (PLoS) in PLOS Computational Biology
- Vol. 21 (8) , e1013371
Abstract
The geometrical and statistical properties of brain activity depend on the way neurons connect to form recurrent circuits. However, the link between connectivity structure and emergent activity remains incompletely understood. We investigate this relationship in recurrent neural networks with additive stochastic inputs. We assume that the synaptic connectivity can be expressed in a low-rank form, parameterized by a handful of connectivity vectors, and examine how the geometry of emergent activity relates to these vectors. Our findings reveal that this relationship critically depends on the dimensionality of the external stochastic inputs. When inputs are low-dimensional, activity remains low-dimensional, and recurrent dynamics influence it within a subspace spanned by a subset of the connectivity vectors, with dimensionality equal to the rank of the connectivity matrix. In contrast, when inputs are high-dimensional, activity also becomes potentially high-dimensional. The contribution of recurrent dynamics is apparent within a subspace spanned by the totality of the connectivity vectors, with dimensionality equal to twice the rank of the connectivity matrix. Applying our formalism to excitatory-inhibitory networks, we discuss how the input configuration also plays a crucial role in determining the amount of amplification generated by non-normal dynamics. Our work provides a foundation for studying activity in structured brain circuits under realistic noise conditions, and offers a framework for interpreting stochastic models inferred from experimental data.Keywords
All Related Versions
Funding Information
- Simons Foundation (00002313)
- Fundação para a Ciência e a Tecnologia (10.54499/2020.05021.BD)
- Simons Foundation (543009)
- Simons Foundation (2794-04)
- NIH Blueprint for Neuroscience Research (R01 EY035896)
- NIH Blueprint for Neuroscience Research (RF1 NS127107)
This publication has 50 references indexed in Scilit:
- Stabilizing machine learning prediction of dynamics: Novel noise-inspired regularization tested with reservoir computingNeural Networks, 2023
- Neural Trajectories in the Supplementary Motor Area and Motor Cortex Exhibit Distinct Geometries, Compatible with Different Classes of ComputationNeuron, 2020
- Cortical Areas Interact through a Communication SubspaceNeuron, 2019
- Circuit Models of Low-Dimensional Shared Variability in Cortical NetworksNeuron, 2019
- Linking Connectivity, Dynamics, and Computations in Low-Rank Recurrent Neural NetworksNeuron, 2018
- Attractor Dynamics in Networks with Learning Rules Inferred from In Vivo DataNeuron, 2018
- Optimal Degrees of Synaptic ConnectivityNeuron, 2017
- On simplicity and complexity in the brave new world of large-scale neuroscienceCurrent Opinion in Neurobiology, 2015
- Balanced Amplification: A New Mechanism of Selective Amplification of Neural Activity PatternsNeuron, 2009
- Mean-field equations, bifurcation map and route to chaos in discrete time neural networksPhysica D: Nonlinear Phenomena, 1994