Appears in collection : 2023 - T1B - WS2 - Networks of spiking neurons

While neurons are receiving discrete synaptic impacts, they are exposed to background fluctuations in the brain, which have been a bottleneck in estimating inter-neuronal connectivity. Though monosynaptic connectivity between neurons can be estimated by capturing tiny interdependence between their firings, spike trains in vivo are originally far from independent, exhibiting large undulation in their cross-correlograms (CCs), and accordingly, classical analysis methods made many false inferences. Modern analysis methods including our GLMCC have overcome the difficulty by smoothing out large undulations of CCs. By analyzing spike trains in more detail, we have newly found another problem that has not been solved even with the modern method. In this talk, I will demonstrate the problem and show how I resolved it.