Efficient supervised learning in networks with binary synapses

I think this paper might fit in with the Numenta view of the neuron:
http://www.pnas.org/content/104/26/11079.full