Building Recurrent Neural Networks to Implement Multiple Attractor Dynamics Using the Gradient Descent Method

<table>Error and Kullback-Leibler divergence between
the teaching sequences and output generated by the RNN for 200 000 learning steps in experiment 1.</table>

Advances in Artificial Neural Systems

fig7

Figure 7

Figure 7: Building Recurrent Neural Networks to Implement Multiple Attractor Dynamics Using the Gradient Descent Method 