The link below is a good review article on the minicolumn hypothesis.

I don't speak for the whole group, but my understanding is that NuPIC is
based on minicolumns as the the fundamental computational unit.

Macrocolumns are essentially structures consisting of many minicolumns.

You're right though, there is no consensus on this question (the
significance cortical columns). There is a reasonable amount of supporting
evidence, but since we don't fully understand how these columns operate so
it remains a hypothesis. As the link you referenced points out, it is
possible that the repetitive structure and organization of these columns
could be the result of patterns of development rather than function. My
feeling is that no doubt development plays a significant role in the
organizational structure, that does not preclude it from establishing a
significant functional role as well (i.e. given the small-world network
organization of the brain, the developmental origins may well have given
rise to the functional capabilities).

In any case, it's fairly evident that the NuPIC community has a good deal
of confidence in the cortical columnar hypothesis being correct.


http://brain.oxfordjournals.org/content/125/5/935.full


On Mon, Apr 14, 2014 at 12:41 PM, mraptor <[email protected]> wrote:

> Hi,
>
> Can somebody point me to a place where I can find more information
> about internal structure of the cortex "minicolumns" and how they
> connect with the neighbouring areas ? If there is such info somewhere
> ?
> It is my understanding the the NuPic is based on the idea of columnar
> structure of the cortex !? right ? Which one do you have in mind
> "minicolumn" OR "macrocolumn" ?
>
> I'm interested for the time being mostly in the minicolumn structure.
>
>
> I'm reading this :
> http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1569491/
>
> and it seems there seems to be some disagreement in "Brain" community
> which one is the basic computational building block of the cortex ...
> minicolumn, modules, macrocolumn .... etc
>
> Can you clarify which part NuPic has picked as basic computational
> artefact ?
>
> thanks..
>
> Again my primary question is any link to info that discusses
> "minicolumns" and in intra/outro-connections ?
>
>
> -------| http://ifni.co
>
> _______________________________________________
> nupic mailing list
> [email protected]
> http://lists.numenta.org/mailman/listinfo/nupic_lists.numenta.org
>
_______________________________________________
nupic mailing list
[email protected]
http://lists.numenta.org/mailman/listinfo/nupic_lists.numenta.org

Reply via email to