Open Access Publication of the Month – Agnes Korcsak-Gorzo (INM-6) et al.

16 May 2022
To encourage us all to broaden our horizons, each month the Central Library selects one open access publication from the JuSER publications portal to be featured in the newsletter of Forschungszentrum Jülich.
In the open access publication of the month, Agnes Korcsak-Gorzo, Michael G. Müller, Andreas Baumbach, Luziwei Leng, Oliver J. Breitwieser, Sacha J. van Albada, Walter Senn, Karlheinz Meier, Robert Legenstein, and Mihai A. Petrovici (a collaboration between Jülich's INM-6, INM-10, and IAS-6 institutes, TU Graz, the University of Bern, and Heidelberg University) present simulation results and a mathematical derivation showing how oscillations can help biologically inspired neural networks to recall information.
In situations where several different interpretations appear plausible, the brain is very good in quickly switching between these interpretations (e.g. is it a bush or a tiger?). This functionality puts the brain ahead of many biologically inspired algorithms when it comes to high-dimensional input such as realistic visual impressions.
The authors took inspiration from this and improved their model using neurobiological phenomena as a paradigm. An example of such phenomenon in the brain are global oscillations – brain waves – that occur in different frequencies and whose function has yet to be fully explained. The authors show that such oscillations help a neural network to recall all plausible interpretations and not take an unnecessarily long time. These theoretical findings indicate a possible function of brain waves. The open access publication entitled “Cortical oscillations support sampling-based computations in spiking neural networks” was published in the journal PLoS Computational Biology.
JuSER publications portal – Cortical oscillations support sampling-based computations in spiking neural networks
Internet – Institute of Neuroscience and Medicine – Computational and Systems Neuroscience (INM-6)