A balanced memory network - Quantitative Biology Neurons and CognitionReport as inadecuate




A balanced memory network - Quantitative Biology Neurons and Cognition - Download this document for free, or read online. Document in PDF available to download.

Abstract: A fundamental problem in neuroscience is understanding how working memory -the ability to store information at intermediate timescales, like 10s ofseconds - is implemented in realistic neuronal networks. The most likelycandidate mechanism is the attractor network, and a great deal of effort hasgone toward investigating it theoretically. Yet, despite almost a quartercentury of intense work, attractor networks are not fully understood. Inparticular, there are still two unanswered questions. First, how is it thatattractor networks exhibit irregular firing, as is observed experimentallyduring working memory tasks? And second, how many memories can be stored underbiologically realistic conditions? Here we answer both questions by studying anattractor neural network in which inhibition and excitation balance each other.Using mean field analysis, we derive a three-variable description of attractornetworks. From this description it follows that irregular firing can exist onlyif the number of neurons involved in a memory is large. The same mean fieldanalysis also shows that the number of memories that can be stored in a networkscales with the number of excitatory connections, a result that has beensuggested for simple models but never shown for realistic ones. Both of thesepredictions are verified using simulations with large networks of spikingneurons.



Author: Yasser Roudi, Peter E. Latham

Source: https://arxiv.org/



DOWNLOAD PDF




Related documents