RE: Internal playback
April 26, 2015 at 9:10 am
(This post was last modified: April 26, 2015 at 9:13 am by emjay.)
(April 26, 2015 at 8:23 am)robvalue Wrote: Wow, that's very interesting. Thanks!
It is so frustrating when I know damn well that I have the information I need in my brain, but I can't find it. I can't do a search function. I can't really do anything, except stop thinking about it. That seems to get me there the quickest. When I become aware I am thinking, I get kind of self conscious and the process starts seeming ridiculous.
Happens to me all the time too

Once you've forgotten the triggers then finding them again is not going to be easy and would take a lot of detective work. The only way I think is with a process of deduction. I would try to think about what would be most likely to 'progress the state' from a neural network point of view.
One thought is to look for something that would be common to many aspects of the memory, i.e. if it's interconnected it will share its activation with the other related aspects and 'prime' them making them easier to activate. But that's not all there is to it because that may not be the initial step: it may share with everything but it might not get them above the firing threshold.
The other assumption I would take for granted and I think would be helpful is that a neural network doesn't treat every input equally - they are weighted. So you could have a whole hierarchy of strongly activated neurons feeding into a what I'll call 'apex' neuron(s) but if that neuron has a small weight in its connection with the next neuron, it probably won't activate it. In other words in this case, of trying to activate a long lost memory, it could be a red herring because although there's a lot of activity it's not considered essential to the memory you're trying to trigger.
I don't know yet how to do it - how to find the most relevant aspects. That's a work in progress but I'm convinced that the answer's there somewhere in neural network mechanisms but just have to find it
