Firstly at fast timescales, information may be stored in the dynamic states of neural activity. Using the newly assembled whole, self-consistent C. elegans connectome and linear systems theory, the propagation of neuronal activity in response to sensory stimulation is explored and several activity patterns that could serve as substrates of previously described behaviors are determined. Secondly at intermediate timescales, information may be stored in the synaptic strengths between mammalian neocortical neurons. Experimental investigations have revealed that synapses possess interesting and, in some cases, unexpected properties. I present a theoretical framework based on maximizing information storage capacity of neural tissue under resource constraints that accounts for three of these properties: typical central synapses are noisy, the distribution of synaptic weights among central synapses is wide, and synaptic connectivity between neurons is sparse. Finally at long timescales, information may be written as symbols on a rewriteable medium. For scenarios when representing information and modifying the representation to update the stored message are both expensive, I describe malleable codes and discuss the fundamental trade off between compression efficiency and malleability cost, measured with a string edit distance; the trade off is demarcated by the solution to a graph embedding problem.

Host: Luis Bettencourt, T-5