Spontaneous Generalization

Connectionist networks may be designed so that they can retrieve information from cues that are too vague to match a particular memory and provide a generalized picture of what is common to the memories that match the cues. Thus the network has the ability to generalize about classes of memories as part of its architecture.

References:

1. Bechtel, W., & Abrahamsen, A. (1991). Connectionism and the mind: An introduction to parallel processing in networks. Cambridge, MA: Blackwell.