Local Representation

A local representation in a connectionist network occurs when a single concept or meaning is represented by the activity of a single processing unit. In other word, the unit responds to the presence of some trigger feature (Barlow, 1972), and so is akin to a "grandmother neuron".

Local representations have the advantage of making the internal structure of artificial neural networks easier to interpret (Dawson, 2004). However, they have the disadvantage of making the system brittle: damage to local representations in a network can result in large behavioral deficits, even though the network is connectionist

References:

1. Barlow, H. B. (1972). Single units and sensation: A neuron doctrine for perceptual psychology? Perception. 1, 371-394.
2. Dawson, M. R. W. (2004). Minds And Machines : Connectionism And Psychological Modeling. Malden, MA: Blackwell Pub.
3. Medler, D. A., Dawson, M. R. W., & Kingstone, A. (2005). Functional localization and double dissociations: The relationship between internal structure and behavior. Brain and Cognition, 57, 146-150.

.(Added November 2009)