Eliasmith, C. et al. Science 338, 1202–1205 (2012).

Simulating how our brains work using super-powerful computers is becoming more and more popular. In further improvement of these models, the guiding theme has been to incorporate the largest possible number of neurons, but relatively little attention has been put on reproducing functionality and behavior. Work by Eliasmith et al. attempts to fill this gap by simulating a variety of complex human behaviors. The model is called Semantic Pointer Architecture Unified Network (or 'Spaun'), and it is based on the activity of 2.5 million simulated neurons that are organized into subsystems resembling different human brain areas and their wiring. When the researchers present Spaun with stimuli, it responds via a physically modeled arm. Spaun can perform diverse tasks involving simple perception to complex cognition, and it may serve to study important aspects of human brain function.