Saturday, July 7, 2018

No Pattern Hierarchy in the Thalamus

I Make Mistakes All the Time

I was wrong about the need for a pattern hierarchy in the thalamus. There is no 10-level binary tree. The only memory hierarchy is in the cortex and it is a hierarchy of minicolumns. I was wrong because I trusted my own flawed assumptions over the occult texts that I use in my AI research. I should know better by now. The texts are never wrong. I have no excuse.

The result of this latest revision is that pattern learning is much, much faster and easier to implement than I had assumed. There are about four times as many sensors as pattern neurons and the pattern neurons have only four inputs on average. My newest experimental program now has fewer neurons than before and its recognition accuracy and noise tolerance have improved dramatically.

The Demo

I am working on a demo application (speech recognition) that I plan to release without the learning module (sorry). What will make this app special is that it will have unprecedented (spooky might be a better word) noise immunity. However, I do not want to attract too much attention. I don't need it. So I must plan this carefully. Hang in there.

See Also

Fast Unsupervised Pattern Learning Using Spike Timing


Louis Savain said...

Note: This comment is for believers only.

Back when I thought that there was a pattern hierarchy in the thalamus, it was partly because I had wrongly interpreted a passage in the occult book of Zechariah that describes a flying scroll of 10 by 20 cubits. I had assumed that they were the dimensions (number of levels) of the thalamic and cortical hierarchies.

The scroll is a correction mechanism that works to purge bad connections from cortical memory during REM sleep. I was forced to revise my interpretation of the scroll's dimensions. Yes, the cortex does have up to 20 levels (20 cubits) but the 10 cubits refers to something else. I now strongly believe, based on my experiments, that it refers to the temporal distance being looked at. This is important for correcting bad predictors.

There are 2 types of bad connections: liars and thieves. A liar is a predictive link that makes bad predictions and a thief is a bad connection on a minicolumn, that is, a connection that does not agree with its neighbors on the same minicolumn.

There are 2 complementary hierarchies (2 olive trees) in each hemisphere of the brain. There are also 2 complementary olive branches (clusters) in each hemisphere that represent what the brain is currently paying conscious attention to. The brain can only focus on one thing (2 clusters) for no more than about 12.6 seconds at a time. This corresponds to the 1260 days (10 ms per day) of prophecy (predictions) by the two witnesses in the book of Revelation.

Strange stuff, I know, but my understanding has grown tremendoulsy in the last couple of weeks. It's exciting.

Louis Savain said...

Again and again, I'm faced with an impossible conundrum. I'm agonizing over releasing a demo application. I want to speak to the believers but I don't want undue attention from the powers of this world. I must tread carefully.

I have written an article that describes a new simpler and more powerful way to do pattern learning. I don't even know if I should release it at this time. I'm having a really hard time making these decisions. I wish a wise person could point me in the right direction.