3
$\begingroup$

As an example of an artificial neural net (ANN), a neural processing unit (NPU) is able to encode previous (learned) information by storing a weighted resistance for each input. Since ANN's are biologically inspired, I was wondering whether the neurons in the human brain also have a mechanism to set the resistance for electrical current that flows through the neuron. If not, how is information (long term storage) encoded in the brain (for example, a familiar word like "house")?

$\endgroup$
4
$\begingroup$

The general biological term for this kind of mechanism is synaptic plasticity. Synapses are the biological structure that allow neurons to communicate via the exchange of neurotransmitters. The strength of the synapse (the effectiveness of the link between the neurons) can be modified over time.

One important kind of modification is long-term potentiation, first reported by Bliss and Lømo in 1973. They demonstrated that short term bursts of activity in a neural pathway can cause long term augmentation of the post-synaptic response. In artificial neural network terms, they were able to demonstrate something akin to an increase in the connection strength between two neurons.

$\endgroup$
  • $\begingroup$ How important is synaptic plasticity when it comes to the information stored in the brain? Does it just assist in the strength of the information encoding or is the plasticity the main encoding mechanism? I think some researchers have been able to get rough pictures of the test subject's visual thoughts - I was always curious how they decoded that information. $\endgroup$ – Wuschelbeutel Kartoffelhuhn May 6 '15 at 16:39
  • $\begingroup$ That could be an interesting new question to ask on the site! $\endgroup$ – Josh de Leeuw May 6 '15 at 16:41

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.