Section: New Results

The associative loop

The prefrontal cortex is known to be involved in many high-level cognitive functions, in particular working memory. Gated working memory is defined as the capacity of holding arbitrary information at any time in order to be used at a later time. Based on electrophysiological recordings, several computational models have tackled the problem using dedicated and explicit mechanisms. We propose instead to consider an implicit mechanism based on a random recurrent neural network. We introduce a robust yet simple reservoir model of gated working memory with instantaneous updates [11]. The model is able to store an arbitrary real value at random time over an extended period of time. The dynamics of the model is a line attractor that learns to exploit reentry and a non-linearity during the training phase using only a few representative values. A deeper study of the model shows that there is actually a large range of hyper parameters for which the results hold (number of neurons, sparsity, global weight scaling, etc.) such that any large enough population, mixing excitatory and inhibitory neurons can quickly learn to realize such gated working memory. This suggests this property could be an implicit property of any random population, that can be acquired through learning. Furthermore, considering working memory to be a physically open but functionally closed system, we give account on some counter-intuitive electrophysiological recordings.

We also developed a model of working memory combining short-term and long-term components [24]. For the long-term component, we used Conceptors in order to store constant temporal patterns. For the short-term component , we used the Gated-Reservoir model [11]. We combined both components in order to obtain a model in which information can go from long-term memory to short-term memory and vice-versa.

The prefrontal cortex is also known to be the place where complex and abstract behavioral rules are implemented. In order to study the mechanisms related to the manipulation of such rules, we have begun the study of networks able to build rules to manipulate such framework as the Wisconsin Card Sorting Test, widely used in the clinical domain [3].