https://www.nature.com/articles/ncomms15276
This paper was published in nature communication in 23.5.2017.
Matthias Ekman is the corresponding author.
HTM is actually the right thing.
In my perspective now,the thinking is that some high level neurons give some input to some other high level neurons and these high level neurons output to other neurons…
the status of neurons constitute the thinking.
anyway,I am so excited about HTM theory and so excited to see that it actually works!
https://arxiv.org/abs/1611.03530
another paper implies that memory is the principle of intelligence …
this paper is not about neuroscience,but this paper analyzed why DL can generalize so well.
In my opinion,these guys who want to achieve intelligence by pure math may fail,because the results show that the more they get close with bioscience the more success they achieve.
from SVM to CNN,it is numenta’s turn now.
A thought experiment from a book of Ray Kurzweil indicates the sequence-memory.
Experiment begins.
Try this:Recite the alphabet.
Now try this:Recite the alphabet backward.
unless you have studied the alphabet in this order,you are likely to find it impossible to do.
end.
book name:how to create a mind.
I’m glad you agree with our approach. 