Larry Abbott – Colombia University
Wednesday, December 7, 10:45AM (Zoom meeting starts at 10:30)
"New Results in Random Network Theory"
I will describe two new results in the theory of randomly connected neural networks. The first is a calculation of the dimension (participation ratio) of the chaotic activity in such networks using a two-site cavity approach. The second involves an expansion of the dynamics of these networks to include time-varying connectivity, which models the ongoing plasticity of the connections (synapses) in biological networks. The dual neuronal-synaptic dynamics in these extended networks reveals some novel phases, including something we call freezable chaos.