<?xml version="1.0" encoding="UTF-8"?><xml><records><record><source-app name="Biblio" version="6.x">Drupal-Biblio</source-app><ref-type>17</ref-type><contributors><authors><author><style face="normal" font="default" size="100%">Sussillo, D.</style></author><author><style face="normal" font="default" size="100%">Abbott, L. D.</style></author></authors></contributors><titles><title><style face="normal" font="default" size="100%">Generating Coherent Patterns of Activity  from Chaotic Neural Networks</style></title><secondary-title><style face="normal" font="default" size="100%">Neuron</style></secondary-title></titles><dates><year><style  face="normal" font="default" size="100%">2009</style></year></dates><volume><style face="normal" font="default" size="100%">63</style></volume><pages><style face="normal" font="default" size="100%">544 - 557</style></pages><abstract><style face="normal" font="default" size="100%">&lt;p&gt;Neural circuits display complex activity patterns both &lt;br /&gt;
spontaneously and when responding to a stimulus or &lt;br /&gt;
generating a motor output. How are these two forms &lt;br /&gt;
of activity related? We develop a procedure called &lt;br /&gt;
FORCE learning for modifying synaptic strengths &lt;br /&gt;
either external to or within a model neural network &lt;br /&gt;
to change chaotic spontaneous activity into a wide &lt;br /&gt;
variety of desired activity patterns. FORCE learning &lt;br /&gt;
works even though the networks we train are sponta- &lt;br /&gt;
neously chaotic and we leave feedback loops intact &lt;br /&gt;
and unclamped during learning. Using this approach, &lt;br /&gt;
we construct networks that produce a wide variety of &lt;br /&gt;
complex output patterns, input-output transforma- &lt;br /&gt;
tions that require memory, multiple outputs that can &lt;br /&gt;
be switched by control inputs, and motor patterns &lt;br /&gt;
matching human motion capture data. Our results &lt;br /&gt;
reproduce data on premovement activity in motor &lt;br /&gt;
and premotor cortex, and suggest that synaptic plas- &lt;br /&gt;
ticity may be a more rapid and powerful modulator of &lt;br /&gt;
network activity than generally appreciated.&lt;/p&gt;</style></abstract></record></records></xml>