Neural network model of memory7/22/2023 As a result, GALIS models can learn both task-specific content and also the necessary cognitive control procedures (instructions) needed to perform a task in the first place. Each region is an attractor network capable of learning temporal sequences, and the individual regions not only exchange task-specific information with each other, but also gate one another's functions and interactions. Models built using GALIS consist of a network of interacting "regions" inspired by the organization of primate cerebral cortex. Here we present a general purpose framework called GALIS that we believe is amenable to developing a broad range of cognitive control models. Developing neurocomputational methods that allow these cognitive control mechanisms to be performed autonomously has proven to be surprisingly difficult. Many neural network models of cognition rely heavily on the modeler for control over aspects of model behavior, such as when to learn and whether an item is judged to be present in memory. We conclude that augmenting simple oscillatory neural network models with temporally asymmetric synaptic connections substantially improves their ability to match human short term memory properties. The network is still capable of matching the recall performance of human subjects, reproducing the recency effect they exhibit in working memory tasks and displaying similar position-specific recall rates. This was achieved through the use of a temporally asymmetric weight matrix. In this paper we modify a recently developed simple oscillatory memory capable of storing temporal sequences so that it will now retrieve remembered items in the same order presented. However, retrieving these stored patterns in the same order as they were seen has proven challenging. ![]() Such oscillatory activity can represent multiple stored patterns simultaneously, rather than the single pattern of a fixed-point network. ![]() Recurrent connections combined with the appropriate dynamics enable oscillatory neural networks to produce rhythmic activity patterns.
0 Comments
Leave a Reply. |