Berlin 2015 – scientific programme
Parts | Days | Selection | Search | Updates | Downloads | Help
BP: Fachverband Biologische Physik
BP 8: Neurophysics II
BP 8.9: Talk
Monday, March 16, 2015, 16:45–17:00, H 1058
Self-consistent spectra in recurrent spiking networks — •Stefan Wieland1,2 and Benjamin Lindner1,2 — 1Bernstein Center for Computational Neuroscience Berlin, Germany — 2Humboldt University Berlin, Germany
Firing patterns in cortical networks are often modeled with Poissonian spike trains. Demanding self-consistency at the level of firing rates, i.e. that spike trains driving a neuron possess the same firing rate as the spike train they evoke, then yields a tractable analytic description of network dynamics. However, output spike trains are usually observed to be non-Poissonian, something a more coherent framework should account for. Here we present iterative schemes that yield self-consistent statistics in recurrent neural networks at the level of spike-train correlations.