. Signatures of neural computation are thought to be reflected in the coordinated activity of large neural populations. Neuroscience is now flush with measurements of population activity in humans, animal subjects, and large-scale artificial network models. In this talk I will address an extensively studied, yet unresolved, question: What statistical methods should we use to quantify whether two or more neural circuits have “similar” population dynamics? Answering this question is of fundamental importance if we are to build theories of cognition that generalize across competing neural circuit models, experimental recording techniques, and model organisms. I will summarize several existing methods to quantify similarity in neural representations, including linear predictivity scores, canonical correlations analysis (CCA), and representational similarity analysis (RSA), and show how they can be unified into a common theoretical framework. I will also show applications of these methods to quantifying individual variability in neural datasets.
Organizer
Devika Narain
d.narain@erasmusmc.nl