ShockLab

Dean Rance

Dean aims to investigate how anatomical homologues of mammalian Social Behaviour Network found in the zebrafish (Danio rerio) brain may regulate social interactions. To do so, he will measure GCaMP fluorescence as a marker of neural activity in zebrafish brains in a social context and using this data, combined with a pair of predictive models of behaviour (conditioned on, and not conditioned on, social cues), he will correlate fluorescence of the anatomical regions with the increased predictability of behaviour when conditioning on social cues. The aim is to identify which anatomical regions determine whether socially-driven behaviours will occur and, thence, determine how said regions may be integrated in a circuit dependent on social cues. The hope is to verify that the zebrafish has a Social Behaviour Network which is functionally similar to the mammalian counterpart, yet which is more amenable to interventional studies.

Masters thesis: Biologically Motivated Reinforcement Learning in Spiking Neural Networks