Our new paper “Differential receptive field organizations give rise to nearly identical neural correlations across three parallel sensory maps in weakly electric fish” was just published in PLoS Computational Biology. Although the journal published it alongside the picture of a skate (which has no connection to the article or its content), I am more than happy with the outcome of the paper. We even were on top of the PLoS Comp Biol page for a few days (see screenshot posted above!). In this paper we approach the question why neural activity in the ELL (electrosensory lateral line lobe) in the medulla of the weakly electric fish is correlated. In electrophysiological recordings we found that, although the spatial tuning properties differ between cells of different sub-areas in the ELL, the level of correlations is surprisingly similar between these.
Using numerical simulations and analytical modeling we explain this phenomenon by the tuning properties of receptive fields. These feedforward contributions differe to differentially tuned center-surround organizations.
As baseline correlations can be seen as the limit towards noise correlations will tend when stimulus amplitude goes to zero, their presence predicts the presence of noise correlations which are known to have a substantial impact on neural coding. Thus our paper gives important insight into the mechanisms by which noise correlations arise in structured feedforward driven neural ensembles.