{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,2,27]],"date-time":"2026-02-27T06:25:12Z","timestamp":1772173512295,"version":"3.50.1"},"update-to":[{"DOI":"10.1371\/journal.pcbi.1013371","type":"new_version","label":"New version","source":"publisher","updated":{"date-parts":[[2025,9,5]],"date-time":"2025-09-05T00:00:00Z","timestamp":1757030400000}}],"reference-count":61,"publisher":"Public Library of Science (PLoS)","issue":"8","license":[{"start":{"date-parts":[[2025,8,18]],"date-time":"2025-08-18T00:00:00Z","timestamp":1755475200000},"content-version":"vor","delay-in-days":0,"URL":"http:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"funder":[{"DOI":"10.13039\/100000893","name":"Simons Foundation","doi-asserted-by":"publisher","award":["00002313"],"award-info":[{"award-number":["00002313"]}],"id":[{"id":"10.13039\/100000893","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/501100001871","name":"Funda\u00e7\u00e3o para a Ci\u00eancia e a Tecnologia","doi-asserted-by":"publisher","award":["10.54499\/2020.05021.BD"],"award-info":[{"award-number":["10.54499\/2020.05021.BD"]}],"id":[{"id":"10.13039\/501100001871","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/100000893","name":"Simons Foundation","doi-asserted-by":"publisher","award":["543009"],"award-info":[{"award-number":["543009"]}],"id":[{"id":"10.13039\/100000893","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/100000893","name":"Simons Foundation","doi-asserted-by":"publisher","award":["2794-04"],"award-info":[{"award-number":["2794-04"]}],"id":[{"id":"10.13039\/100000893","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/100000135","name":"NIH Blueprint for Neuroscience Research","doi-asserted-by":"publisher","award":["R01 EY035896"],"award-info":[{"award-number":["R01 EY035896"]}],"id":[{"id":"10.13039\/100000135","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/100000135","name":"NIH Blueprint for Neuroscience Research","doi-asserted-by":"publisher","award":["RF1 NS127107"],"award-info":[{"award-number":["RF1 NS127107"]}],"id":[{"id":"10.13039\/100000135","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":["www.ploscompbiol.org"],"crossmark-restriction":false},"short-container-title":["PLoS Comput Biol"],"abstract":"<jats:p>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.<\/jats:p>","DOI":"10.1371\/journal.pcbi.1013371","type":"journal-article","created":{"date-parts":[[2025,8,18]],"date-time":"2025-08-18T18:07:57Z","timestamp":1755540477000},"page":"e1013371","update-policy":"https:\/\/doi.org\/10.1371\/journal.pcbi.corrections_policy","source":"Crossref","is-referenced-by-count":0,"title":["Stochastic activity in low-rank recurrent neural networks"],"prefix":"10.1371","volume":"21","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-7682-5178","authenticated-orcid":true,"given":"Francesca","family":"Mastrogiuseppe","sequence":"first","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Joana","family":"Carmona","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Christian K.","family":"Machens","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"340","published-online":{"date-parts":[[2025,8,18]]},"reference":[{"issue":"12","key":"pcbi.1013371.ref001","doi-asserted-by":"crossref","first-page":"1","DOI":"10.1371\/journal.pcbi.1006309","article-title":"Coherent chaos in a recurrent neural network with structured connectivity","volume":"14","author":"I Landau","year":"2018","journal-title":"PLOS Comput Biol"},{"issue":"2","key":"pcbi.1013371.ref002","doi-asserted-by":"crossref","DOI":"10.1016\/j.neuron.2018.11.034","article-title":"Circuit models of low-dimensional shared variability in cortical networks","volume":"101","author":"C Huang","year":"2019","journal-title":"Neuron"},{"key":"pcbi.1013371.ref003","doi-asserted-by":"crossref","DOI":"10.7554\/eLife.68422","article-title":"Global organization of neuronal activity only requires unstructured local connectivity","volume":"11","author":"D Dahmen","year":"2022","journal-title":"Elife"},{"key":"pcbi.1013371.ref004","doi-asserted-by":"crossref","first-page":"148","DOI":"10.1016\/j.conb.2015.04.003","article-title":"On simplicity and complexity in the brave new world of large-scale neuroscience","volume":"32","author":"P Gao","year":"2015","journal-title":"Curr Opin Neurobiol"},{"issue":"1","key":"pcbi.1013371.ref005","doi-asserted-by":"crossref","first-page":"102","DOI":"10.1038\/s41593-017-0028-6","article-title":"Flexible timing by temporal scaling of cortical responses","volume":"21","author":"J Wang","year":"2018","journal-title":"Nat Neurosci"},{"issue":"37","key":"pcbi.1013371.ref006","doi-asserted-by":"crossref","first-page":"23021","DOI":"10.1073\/pnas.1915984117","article-title":"Low-dimensional dynamics for working memory and time encoding","volume":"117","author":"CJ Cueva","year":"2020","journal-title":"Proc Natl Acad Sci U S A"},{"key":"pcbi.1013371.ref007","unstructured":"Eliasmith C, Anderson C. Neural engineering - computation, representation, and dynamics in neurobiological systems. MIT Press; 2004."},{"issue":"3","key":"pcbi.1013371.ref008","doi-asserted-by":"crossref","DOI":"10.1016\/j.neuron.2018.07.003","article-title":"Linking connectivity, dynamics, and computations in low-rank recurrent neural networks","volume":"99","author":"F Mastrogiuseppe","year":"2018","journal-title":"Neuron"},{"issue":"5","key":"pcbi.1013371.ref009","doi-asserted-by":"crossref","first-page":"803","DOI":"10.1162\/neco_a_01658","article-title":"Approximating nonlinear functions with latent boundaries in low-rank excitatory-inhibitory spiking networks","volume":"36","author":"WF Podlaski","year":"2024","journal-title":"Neural Comput"},{"key":"pcbi.1013371.ref010","article-title":"High-dimensional dynamics in low-dimensional networks","author":"Y Wan","year":"2025","journal-title":"arXiv preprint"},{"issue":"1","key":"pcbi.1013371.ref011","doi-asserted-by":"crossref","DOI":"10.1016\/j.neuron.2018.05.038","article-title":"Attractor dynamics in networks with learning rules inferred from in vivo data","volume":"99","author":"U Pereira","year":"2018","journal-title":"Neuron"},{"issue":"2","key":"pcbi.1013371.ref012","doi-asserted-by":"crossref","DOI":"10.1371\/journal.pcbi.1007655","article-title":"Coding with transient trajectories in recurrent neural networks","volume":"16","author":"G Bondanelli","year":"2020","journal-title":"PLoS Comput Biol"},{"issue":"6","key":"pcbi.1013371.ref013","doi-asserted-by":"crossref","first-page":"1572","DOI":"10.1162\/neco_a_01381","article-title":"Shaping dynamics with multiple populations in low-rank recurrent networks","volume":"33","author":"M Beiran","year":"2021","journal-title":"Neural Comput"},{"key":"pcbi.1013371.ref014","doi-asserted-by":"crossref","DOI":"10.7554\/eLife.79908","article-title":"Evolution of neural activity in circuits bridging sensory and abstract knowledge","volume":"12","author":"F Mastrogiuseppe","year":"2023","journal-title":"Elife"},{"key":"pcbi.1013371.ref015","doi-asserted-by":"crossref","first-page":"013176","DOI":"10.1103\/PhysRevResearch.3.013176","article-title":"Quality of internal representation shapes learning performance in feedback neural networks","volume":"3","author":"L Susman","year":"2021","journal-title":"Phys Rev Res"},{"key":"pcbi.1013371.ref016","doi-asserted-by":"crossref","first-page":"011909","DOI":"10.1103\/PhysRevE.86.011909","article-title":"Non-normal amplification in random balanced neuronal networks","volume":"86","author":"G Hennequin","year":"2012","journal-title":"Phys Rev E Stat Nonlin Soft Matter Phys"},{"key":"pcbi.1013371.ref017","doi-asserted-by":"crossref","DOI":"10.7554\/eLife.23978","article-title":"Attentional modulation of neuronal variability in circuit models of cortex","volume":"6","author":"T Kanashiro","year":"2017","journal-title":"Elife"},{"issue":"7","key":"pcbi.1013371.ref018","article-title":"The spectrum of covariance matrices of randomly connected recurrent neuronal networks with linear dynamics","volume":"18","author":"Y Hu","year":"2022","journal-title":"PLoS Comput Biol"},{"issue":"1","key":"pcbi.1013371.ref019","doi-asserted-by":"crossref","first-page":"334","DOI":"10.1523\/JNEUROSCI.13-01-00334.1993","article-title":"The highly irregular firing of cortical cells is inconsistent with temporal integration of random epsps","volume":"13","author":"WR Softky","year":"1993","journal-title":"J Neurosci"},{"issue":"10","key":"pcbi.1013371.ref020","doi-asserted-by":"crossref","first-page":"3870","DOI":"10.1523\/JNEUROSCI.18-10-03870.1998","article-title":"The variable discharge of cortical neurons: implications for connectivity, computation, and information coding","volume":"18","author":"MN Shadlen","year":"1998","journal-title":"J Neurosci"},{"issue":"7474","key":"pcbi.1013371.ref021","doi-asserted-by":"crossref","first-page":"78","DOI":"10.1038\/nature12742","article-title":"Context-dependent computation by recurrent dynamics in prefrontal cortex","volume":"503","author":"V Mante","year":"2013","journal-title":"Nature"},{"issue":"2","key":"pcbi.1013371.ref022","article-title":"Training excitatory-inhibitory recurrent neural networks for cognitive tasks: a simple and flexible framework","volume":"12","author":"HF Song","year":"2016","journal-title":"PLoS Comput Biol"},{"key":"pcbi.1013371.ref023","unstructured":"Jaeger H. Adaptive nonlinear system identification with echo state networks. In: NIPS. 2002. p. 609\u201316."},{"key":"pcbi.1013371.ref024","doi-asserted-by":"crossref","first-page":"94","DOI":"10.1016\/j.neunet.2023.10.054","article-title":"Stabilizing machine learning prediction of dynamics: novel noise-inspired regularization tested with reservoir computing","volume":"170","author":"A Wikner","year":"2024","journal-title":"Neural Netw"},{"key":"pcbi.1013371.ref025","doi-asserted-by":"crossref","DOI":"10.7554\/eLife.93060","article-title":"Aligned and oblique dynamics in recurrent neural networks","volume":"13","author":"F Schuessler","year":"2024","journal-title":"Elife"},{"key":"pcbi.1013371.ref026","first-page":"13352","article-title":"The interplay between randomness and structure during learning in RNNs","volume":"33","author":"F Schuessler","year":"2020","journal-title":"Adv Neural Inf Process Syst"},{"key":"pcbi.1013371.ref027","article-title":"Dynamically learning to integrate in recurrent neural networks","author":"B Bordelon","year":"2025","journal-title":"arXiv preprint"},{"key":"pcbi.1013371.ref028","unstructured":"Ghahramani Z, Hinton G. Parameter estimation for linear dynamical systems. Technical Report CRG-TR-96-2; 1996."},{"issue":"5","key":"pcbi.1013371.ref029","doi-asserted-by":"crossref","first-page":"965","DOI":"10.1162\/089976603765202622","article-title":"Estimating a state-space model from point process observations","volume":"15","author":"AC Smith","year":"2003","journal-title":"Neural Comput"},{"key":"pcbi.1013371.ref030","article-title":"Inferring stochastic low-rank recurrent neural networks from neural data","volume":"37","author":"M Pals","year":"2024","journal-title":"Adv Neural Inf Process Syst"},{"issue":"8","key":"pcbi.1013371.ref031","doi-asserted-by":"crossref","first-page":"2554","DOI":"10.1073\/pnas.79.8.2554","article-title":"Neural networks and physical systems with emergent collective computational abilities","volume":"79","author":"JJ Hopfield","year":"1982","journal-title":"Proc Natl Acad Sci USA"},{"key":"pcbi.1013371.ref032","unstructured":"H. Risken. The Fokker-Planck equation. Springer; 2012."},{"issue":"6","key":"pcbi.1013371.ref033","doi-asserted-by":"crossref","first-page":"783","DOI":"10.1038\/s41593-022-01088-4","article-title":"The role of population structure in computations through neural dynamics","volume":"25","author":"A Dubreuil","year":"2022","journal-title":"Nat Neurosci"},{"key":"pcbi.1013371.ref034","article-title":"Extracting computational mechanisms from neural data using low-rank rnns","volume":"35","author":"A Valente","year":"2022","journal-title":"Adv Neural Inf Process Syst"},{"key":"pcbi.1013371.ref035","unstructured":"Yule G. Statistical study of literary vocabulary. Cambridge Press; 1944."},{"issue":"5","key":"pcbi.1013371.ref036","doi-asserted-by":"crossref","DOI":"10.1016\/j.neuron.2017.01.030","article-title":"Optimal degrees of synaptic connectivity","volume":"93","author":"A Litwin-Kumar","year":"2017","journal-title":"Neuron"},{"key":"pcbi.1013371.ref037","article-title":"The low-rank eigenvalue problem","author":"Y Nakatsukasa","year":"2019","journal-title":"arXiv preprint"},{"issue":"1","key":"pcbi.1013371.ref038","doi-asserted-by":"crossref","first-page":"24","DOI":"10.1016\/0167-2789(94)90024-8","article-title":"Mean-field equations, bifurcation map and route to chaos in discrete time neural networks","volume":"74","author":"B Cessac","year":"1994","journal-title":"Physica D"},{"key":"pcbi.1013371.ref039","article-title":"Sequence to sequence learning with neural networks","volume":"27","author":"I Sutskever","year":"2014","journal-title":"Adv Neural Inf Process Syst"},{"issue":"7765","key":"pcbi.1013371.ref040","doi-asserted-by":"crossref","first-page":"361","DOI":"10.1038\/s41586-019-1346-5","article-title":"High-dimensional geometry of population responses in visual cortex","volume":"571","author":"C Stringer","year":"2019","journal-title":"Nature"},{"issue":"11","key":"pcbi.1013371.ref041","doi-asserted-by":"crossref","first-page":"1","DOI":"10.1371\/journal.pcbi.1003258","article-title":"Predictive coding of dynamical variables in balanced spiking networks","volume":"9","author":"M Boerlin","year":"2013","journal-title":"PLOS Comput Biol"},{"key":"pcbi.1013371.ref042","article-title":"Predictive coding in balanced neural networks with noise, chaos and delays","volume":"33","author":"J Kadmon","year":"2020","journal-title":"Adv Neural Inf Process Syst"},{"issue":"23","key":"pcbi.1013371.ref043","doi-asserted-by":"crossref","first-page":"13339","DOI":"10.1073\/pnas.93.23.13339","article-title":"How the brain keeps the eyes still","volume":"93","author":"HS Seung","year":"1996","journal-title":"Proc Natl Acad Sci U S A"},{"issue":"4","key":"pcbi.1013371.ref044","doi-asserted-by":"crossref","first-page":"635","DOI":"10.1016\/j.neuron.2009.02.005","article-title":"Balanced amplification: a new mechanism of selective amplification of neural activity patterns","volume":"61","author":"BK Murphy","year":"2009","journal-title":"Neuron"},{"key":"pcbi.1013371.ref045","doi-asserted-by":"crossref","first-page":"2025","DOI":"10.1175\/1520-0469(1996)053<2025:GSTPIA>2.0.CO;2","article-title":"Generalized stability theory. Part I: autonomous operators","volume":"53","author":"B Farrell","year":"1996","journal-title":"J Atmos Sci"},{"issue":"11","key":"pcbi.1013371.ref046","doi-asserted-by":"crossref","first-page":"683","DOI":"10.1038\/s41583-022-00634-0","article-title":"Multiregion neuronal activity: the forest and the trees","volume":"23","author":"TA Machado","year":"2022","journal-title":"Nat Rev Neurosci"},{"issue":"7495","key":"pcbi.1013371.ref047","doi-asserted-by":"crossref","first-page":"207","DOI":"10.1038\/nature13186","article-title":"A mesoscale connectome of the mouse brain","volume":"508","author":"SW Oh","year":"2014","journal-title":"Nature"},{"issue":"7879","key":"pcbi.1013371.ref048","doi-asserted-by":"crossref","first-page":"159","DOI":"10.1038\/s41586-021-03970-w","article-title":"Cellular anatomy of the mouse primary motor cortex","volume":"598","author":"R Mu\u00f1oz-Casta\u00f1eda","year":"2021","journal-title":"Nature"},{"key":"pcbi.1013371.ref049","article-title":"Partial observation can induce mechanistic mismatches in data-constrained models of neural dynamics","volume":"38","author":"W Quian","year":"2024","journal-title":"Adv Neural Inf Process Syst"},{"key":"pcbi.1013371.ref050","article-title":"Direct neural perturbations reveal a dynamical mechanism for robust computation","author":"D O\u2019Shea","year":"2022","journal-title":"bioRxiv"},{"key":"pcbi.1013371.ref051","doi-asserted-by":"crossref","unstructured":"Kumar A, Frank L, Bouchard K. Identifying feedforward and feedback controllable subspaces of neural population dynamics. 2024. https:\/\/arxiv.org\/abs\/2408.05875","DOI":"10.21203\/rs.3.rs-4102129\/v1"},{"issue":"18","key":"pcbi.1013371.ref052","doi-asserted-by":"crossref","first-page":"188104","DOI":"10.1103\/PhysRevLett.97.188104","article-title":"Eigenvalue spectra of random matrices for neural networks","volume":"97","author":"K Rajan","year":"2006","journal-title":"Phys Rev Lett"},{"issue":"4","key":"pcbi.1013371.ref053","doi-asserted-by":"crossref","DOI":"10.1016\/j.neuron.2020.05.020","article-title":"Neural trajectories in the supplementary motor area and motor cortex exhibit distinct geometries, compatible with different classes of computation","volume":"107","author":"AA Russo","year":"2020","journal-title":"Neuron"},{"issue":"3","key":"pcbi.1013371.ref054","doi-asserted-by":"crossref","first-page":"626","DOI":"10.1162\/NECO_a_00409","article-title":"Opening the black box: low-dimensional dynamics in high-dimensional recurrent neural networks","volume":"25","author":"D Sussillo","year":"2013","journal-title":"Neural Comput"},{"key":"pcbi.1013371.ref055","doi-asserted-by":"crossref","DOI":"10.7554\/eLife.54875","article-title":"Inhibition stabilization is a widespread property of cortical networks","volume":"9","author":"A Sanzeni","year":"2020","journal-title":"Elife"},{"issue":"1","key":"pcbi.1013371.ref056","doi-asserted-by":"crossref","first-page":"107","DOI":"10.1038\/nn.4433","article-title":"The spatial structure of correlated neuronal variability","volume":"20","author":"R Rosenbaum","year":"2017","journal-title":"Nat Neurosci"},{"issue":"15","key":"pcbi.1013371.ref057","doi-asserted-by":"crossref","DOI":"10.1073\/pnas.1802356115","article-title":"Nonmonotonic spatial structure of interneuronal correlations in prefrontal microcircuits","volume":"115","author":"S Safavi","year":"2018","journal-title":"Proc Natl Acad Sci U S A"},{"issue":"1","key":"pcbi.1013371.ref058","doi-asserted-by":"crossref","first-page":"44","DOI":"10.1038\/s41467-021-27724-4","article-title":"Cortical state dynamics and selective attention define the spatial pattern of correlated variability in neocortex","volume":"13","author":"Y-L Shi","year":"2022","journal-title":"Nat Commun"},{"issue":"10","key":"pcbi.1013371.ref059","doi-asserted-by":"crossref","DOI":"10.1073\/pnas.2404039122","article-title":"Structure of activity in multiregion recurrent neural networks","volume":"122","author":"DG Clark","year":"2025","journal-title":"Proc Natl Acad Sci U S A"},{"issue":"1","key":"pcbi.1013371.ref060","doi-asserted-by":"crossref","first-page":"6837","DOI":"10.1038\/s41467-023-42519-5","article-title":"Early selection of task-relevant features through population gating","volume":"14","author":"J Barbosa","year":"2023","journal-title":"Nat Commun"},{"issue":"1","key":"pcbi.1013371.ref061","doi-asserted-by":"crossref","first-page":"249","DOI":"10.1016\/j.neuron.2019.01.026","article-title":"Cortical areas interact through a communication subspace","volume":"102","author":"J Semedo","year":"2019","journal-title":"Neuron"}],"updated-by":[{"DOI":"10.1371\/journal.pcbi.1013371","type":"new_version","label":"New version","source":"publisher","updated":{"date-parts":[[2025,9,5]],"date-time":"2025-09-05T00:00:00Z","timestamp":1757030400000}}],"container-title":["PLOS Computational Biology"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/dx.plos.org\/10.1371\/journal.pcbi.1013371","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,9,5]],"date-time":"2025-09-05T18:07:58Z","timestamp":1757095678000},"score":1,"resource":{"primary":{"URL":"https:\/\/dx.plos.org\/10.1371\/journal.pcbi.1013371"}},"subtitle":[],"editor":[{"given":"Oren","family":"Shriki","sequence":"first","affiliation":[],"role":[{"role":"editor","vocabulary":"crossref"}]}],"short-title":[],"issued":{"date-parts":[[2025,8,18]]},"references-count":61,"journal-issue":{"issue":"8","published-online":{"date-parts":[[2025,8,18]]}},"URL":"https:\/\/doi.org\/10.1371\/journal.pcbi.1013371","relation":{"has-preprint":[{"id-type":"doi","id":"10.1101\/2025.04.22.649933","asserted-by":"object"}]},"ISSN":["1553-7358"],"issn-type":[{"value":"1553-7358","type":"electronic"}],"subject":[],"published":{"date-parts":[[2025,8,18]]}}}