000890946 001__ 890946
000890946 005__ 20240313103115.0
000890946 0247_ $$2Handle$$a2128/28511
000890946 037__ $$aFZJ-2021-01268
000890946 1001_ $$0P:(DE-Juel1)171384$$aKeup, Christian$$b0$$eCorresponding author$$ufzj
000890946 1112_ $$aCOSYNE 2021$$cvirtual$$d2021-02-24 - 2021-02-26$$wNA
000890946 245__ $$aTransient chaotic SNR amplification
000890946 260__ $$c2021
000890946 3367_ $$033$$2EndNote$$aConference Paper
000890946 3367_ $$2BibTeX$$aINPROCEEDINGS
000890946 3367_ $$2DRIVER$$aconferenceObject
000890946 3367_ $$2ORCID$$aCONFERENCE_POSTER
000890946 3367_ $$2DataCite$$aOutput Types/Conference Poster
000890946 3367_ $$0PUB:(DE-HGF)24$$2PUB:(DE-HGF)$$aPoster$$bposter$$mposter$$s1629116318_5910$$xAfter Call
000890946 500__ $$aThe short summary video can be found online at https://vimeo.com/514398744
000890946 520__ $$aStrongly chaotic non-linear networks strongly separate inputs, but are believed to be useless for classification tasksbecause also irrelevant (noise) differences within any class are exacerbated, leading to bad generalization. We show thisis actually not the case during the initial time period following input presentation: During this time, the representationis dominated by expansion, but not by mixing, and larger differences (between classes) expand faster than smallerdifferences (within classes). Therefore, the representation is disentangled by the dynamics, and when classifying thenetwork state by linear readouts, the signal-to-noise ratio (SNR) actually increases, before it eventually deteriorateswhen mixing begins to dominate. We show that this is a general effect in high-dimensional non-linear chaotic systems,and demonstrate it in spiking, continuous rate, and LSTM networks. The transient SNR amplification is always fast(within 50 ms) for spiking networks, while its timescale in continuous valued networks depends on the distance to theedge of chaos. Moreover, this fast, noise-resilient transient disentanglement of representations is in line with empiricalevidence: the olfactory bulb, for example, rapidly enhances the separability of sensory representations in a singlerecurrent layer, being the initial processing stage of a relatively flat hierarchy.
000890946 536__ $$0G:(DE-HGF)POF4-5232$$a5232 - Computational Principles (POF4-523)$$cPOF4-523$$fPOF IV$$x0
000890946 536__ $$0G:(DE-82)EXS-SF-neuroIC002$$aneuroIC002 - Recurrence and stochasticity for neuro-inspired computation (EXS-SF-neuroIC002)$$cEXS-SF-neuroIC002$$x1
000890946 536__ $$0G:(DE-Juel1)HGF-SMHB-2014-2018$$aMSNN - Theory of multi-scale neuronal networks (HGF-SMHB-2014-2018)$$cHGF-SMHB-2014-2018$$fMSNN$$x2
000890946 536__ $$0G:(GEPRIS)368482240$$aGRK 2416 - GRK 2416: MultiSenses-MultiScales: Neue Ansätze zur Aufklärung neuronaler multisensorischer Integration (368482240)$$c368482240$$x3
000890946 536__ $$0G:(DE-82)EXS-PF$$aERS Prep Fund - Exploratory Research Space: Prep Fund als Anschubfinanzierung zur Schließung strategischer Lücken (EXS-PF)$$cEXS-PF$$x4
000890946 7001_ $$0P:(DE-Juel1)164473$$aKühn, Tobias$$b1
000890946 7001_ $$0P:(DE-Juel1)156459$$aDahmen, David$$b2$$ufzj
000890946 7001_ $$0P:(DE-Juel1)144806$$aHelias, Moritz$$b3$$eLast author$$ufzj
000890946 8564_ $$uhttps://juser.fz-juelich.de/record/890946/files/%285min_summary%29_Transient_chaotic_SNR_amplification.mp4$$yRestricted
000890946 8564_ $$uhttps://juser.fz-juelich.de/record/890946/files/%28extended_abstract%29_Transient_chaotic_SNR_amplification.pdf$$yOpenAccess
000890946 8564_ $$uhttps://juser.fz-juelich.de/record/890946/files/%28poster_slides%29_Transient_chaotic_SNR_amplification.pdf$$yRestricted
000890946 909CO $$ooai:juser.fz-juelich.de:890946$$pdriver$$pVDB$$popen_access$$popenaire
000890946 9101_ $$0I:(DE-588b)5008462-8$$6P:(DE-Juel1)171384$$aForschungszentrum Jülich$$b0$$kFZJ
000890946 9101_ $$0I:(DE-588b)5008462-8$$6P:(DE-Juel1)156459$$aForschungszentrum Jülich$$b2$$kFZJ
000890946 9101_ $$0I:(DE-588b)5008462-8$$6P:(DE-Juel1)144806$$aForschungszentrum Jülich$$b3$$kFZJ
000890946 9131_ $$0G:(DE-HGF)POF4-523$$1G:(DE-HGF)POF4-520$$2G:(DE-HGF)POF4-500$$3G:(DE-HGF)POF4$$4G:(DE-HGF)POF$$9G:(DE-HGF)POF4-5232$$aDE-HGF$$bKey Technologies$$lNatural, Artificial and Cognitive Information Processing$$vNeuromorphic Computing and Network Dynamics$$x0
000890946 9130_ $$0G:(DE-HGF)POF3-574$$1G:(DE-HGF)POF3-570$$2G:(DE-HGF)POF3-500$$3G:(DE-HGF)POF3$$4G:(DE-HGF)POF$$aDE-HGF$$bKey Technologies$$lDecoding the Human Brain$$vTheory, modelling and simulation$$x0
000890946 9141_ $$y2021
000890946 915__ $$0StatID:(DE-HGF)0510$$2StatID$$aOpenAccess
000890946 920__ $$lno
000890946 9201_ $$0I:(DE-Juel1)INM-6-20090406$$kINM-6$$lComputational and Systems Neuroscience$$x0
000890946 9201_ $$0I:(DE-Juel1)IAS-6-20130828$$kIAS-6$$lTheoretical Neuroscience$$x1
000890946 9201_ $$0I:(DE-Juel1)INM-10-20170113$$kINM-10$$lJara-Institut Brain structure-function relationships$$x2
000890946 9801_ $$aFullTexts
000890946 980__ $$aposter
000890946 980__ $$aVDB
000890946 980__ $$aUNRESTRICTED
000890946 980__ $$aI:(DE-Juel1)INM-6-20090406
000890946 980__ $$aI:(DE-Juel1)IAS-6-20130828
000890946 980__ $$aI:(DE-Juel1)INM-10-20170113
000890946 981__ $$aI:(DE-Juel1)IAS-6-20130828