Sequences of neuronal activity in attractor network models
by
Henrik Lindén(CB/CSC/KTH)
→
Europe/Stockholm
RB35
RB35
Description
Many cognitive and motor functions rely on the representation of sensory and internal events over time but is far from understood how neuronal circuits encode sequences of information in a stable fashion. We study a class of attractor memory network in which each memory is stored in a distributed fashion, represented by increased firing in a set of pools of excitatory neurons. Excitatory activity is locally modulated by surrounding inhibitory neurons that thereby generate a type of winner-take-all mechanism. Networks of this type has previously been shown to exhibit switching between low-rate asynchronous activity and high-rate attractor state activations (Lundqvist 2010). This type of memory network is however not capable of storing sequences of information, as there is no reliable association between different attractor states.
Here we extend the previous class of attractor network models with specific synaptic projections that associate distinct network states into sequences of neuronal activity, through a combination of fast (AMPA type) and slower (NMDA type) synapses. By combining a mean-field description and simulations of networks of integrate-and-fire neurons we study how the ability to encode and replay sequences of activity depend on the specific structure of the inhibitory microcircuitry and the local balance of excitation and inhibition in the network. Preliminary results show that the network can reliable store spatiotemporal patterns lasting several seconds in networks of just a few thousand neurons. Moreover, excitatory pools can participate multiple times in the sequence, suggesting that neuronal networks of this type can underly a combinatorial code in which information at one step of the sequence is represented by the combination of active excitatory pools.
Lundqvist, M., Compte, A., & Lansner, A. (2010). Bistable, irregular firing and population oscillations in a modular attractor memory network. PLoS Computational Biology, 6(6)
