Shared publicly  - 
Scrapping Excessive Neural Connections Helps Build New Connections

Full article at

The formation and refinement of neural networks is known to be often an activity-dependent process, but mechanisms and nature of activity are not yet clearly understood. During neuronal circuit formation, afferent axons often make multiple, exuberant, or aberrant connections, which are later eliminated, sometimes in an activity-dependent manner. Such examples are widely seen throughout the central and peripheral nervous systems.

The research is in European Journal of Neuroscience. (full access paywall)

Mari Christian's profile photoLuiz Alberto Luz de Almeida's profile photo
Pruning seems to be a universal stimulant to strong regrowth.
Developing a Spiking Neural Network VO2 Brain

An additional advantage of the array of neurons of VO2 is to behave like optogenetic sensors and actuators for neuronal activities.
Rafael Yuste (Columbia University) is trying optogenetics in hopes of gaining a better understanding of how the brain functions. In my opinion, we can also think of as an electrical inverse problem to study, for example, 100 neurons, using similar techniques to electromagnetic geophysics, to try to understand the function of a worm or a matrix of rat neurons. We can train the VO2 neurons to behave like them, and the multiple realizability thesis may help us to understand the biological brain to improve neuromorfic bains and vice-versa.

Mancuso, J. J., Kim, J., Lee, S., Tsuda, S., Chow, N. B., & Augustine, G. J. (2011). Optogenetic probing of functional brain circuitry. Experimental physiology, 96(1), 26-33.

Rafael Yuste: Imaging of neuro circuits key to understanding brain function

Rat Neurons Grown On A Computer Chip Fly A Simulated Aircraft      - Rat Neurons Grown On A Computer Chip Fly A Simulated Aircraft

Stanford Encyclopedia of Philosophy
Chapter Multiple Realizability
Add a comment...