Dresden 2026 – wissenschaftliches Programm
Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe
DY: Fachverband Dynamik und Statistische Physik
DY 33: Statistical Physics of Biological Systems I (joint session DY/BP)
DY 33.7: Vortrag
Mittwoch, 11. März 2026, 11:15–11:30, ZEU/0114
Improving neuronal information transmission with pathway splitting — •Kolja Klett1,2 and Benjamin Lindner1,2 — 1Humboldt University, Berlin — 2Bernstein Center for Computational Neuroscience, Berlin
In many organisms sensory information can take different neuronal paths from sensory cells to destinations in the brain. These paths can be made up of neurons serving distinct functions. Often, pathways consisting of neurons coding the increase (ON) or decrease (OFF) of a signal are observed which have be found to improve the transmission static signals. Here, we consider a simple network of spiking ON and OFF neurons to study the effects of pathway splitting on the transmission of dynamic signals. To that end, we use the coherence function as a frequency-resolved measure of informations transmission. We relate the information transmission of the whole network to that of the constituting neurons by employing response theory leading to approximate relations for the coherence function. For a simple white noise driven integrate-and-fire model of spiking neurons, we find an optimal mixture of ON and OFF neurons which maximizes the coherence function over a broad frequency range. The effect can be attributed to the nonlinear response of the neurons that only becomes relevant for sufficiently strong stimuli.
Keywords: spiking neuron models; neural information transmission; response theory; neural network