×
Login Register an account
Top Submissions Explore Upgoat Search Random Subverse Random Post Colorize! Site Rules Donate
6

Fruitfly brain mapped.

submitted by Crackinjokes to technology 7 monthsOct 2, 2024 17:00:43 ago (+6/-0)     (archive.md)

https://archive.md/vBUjt

The thing I want to know is could you map this to an llm.


9 comments block


[ - ] BoozyB 1 point 7 monthsOct 2, 2024 18:43:04 ago (+1/-0)

Which one of us Goats was used as a control specimen?

[ - ] Crackinjokes [op] 0 points 7 monthsOct 3, 2024 12:21:08 ago (+0/-0)*

Actually just thinking about it it's only 140,000 nuerons.

So you could literally paste those neurons as cells in a spreadsheet and then have some macro to make the proper connections between cells to match that brain map and then run the whole thing as a spreadsheet.

Microsoft Excell limit is

1,048,576 rows by 16,384 columns

So literally 10 rows of 14,000 columns.

You wouldn't even be near the excel limit.

Here in 2009 they were already simulating brains of nuerons of this size on gpus.

https://www.researchgate.net/figure/Speedup-of-GPU-over-CPU-for-a-network-with-100K-neurons-and-varying-number-of-synaptic_fig3_221534151

[ - ] puremadness 0 points 7 monthsOct 2, 2024 20:27:39 ago (+0/-0)*

What do we learn and how do we learn it.. from this particular output?

I'm certain it has uses and contains useful information.
I'm curious what exactly the "connectome" displays and how I can decipher that, or rather, how the creators of it intend to use it to solve what problems exactly?

I'm asking.. how is this useful. I dont doubt that it will be, just curious how.

The picture is cool.

When I look at that image and read that it is taken from the tiny grub, it makes me think "Did a creator put that much effort into the layout of a maggots mind" and then it makes me wonder if that is the correct question to be asking or if I am witnessing the simple marvel of this natural world, which would then still be a creator, though not anthro.

Either way it's pretty cool and I stand amazed at the tiny thing that I can't explain.

[ - ] deleted 0 points 7 monthsOct 3, 2024 01:56:50 ago (+0/-0)

deleted

[ - ] puremadness 1 point 7 monthsOct 3, 2024 02:11:53 ago (+1/-0)

Yes, that is what I am interested in. The how of that process would be very interesting to me.
Primarily the techniques used to correlate between the data sets (the map and the behavior).
It's true I have 0 understanding of certain statistical systems that are likely in play, but I wanna know.

either way, very cool, I like pondering intricate natural designs, its like the epitome of beauty.

[ - ] deleted 0 points 7 monthsOct 3, 2024 23:37:43 ago (+0/-0)

deleted

[ - ] Crackinjokes [op] 0 points 7 monthsOct 3, 2024 12:18:44 ago (+0/-0)

If you replicated every connection in an ai model and put similar inputs and outputs to reality it would be interesting to track responses.

[ - ] puremadness 0 points 7 monthsOct 3, 2024 15:34:45 ago (+0/-0)

it would, but thats not as simple as it sounds, so I was attempting to demystify that.
It amazes me that people say "well just do this" but that "just do" is fantastically difficult and we all just act as if its already done. Not criticizing your reply, if I critique anything it is the manner in which those people write. I dont like the laymans explanation. I want them to show me the money. I wanna see the process or I might not believe they can do it and might suspect the research will not be used, but simply monetized by bedazzling funders.

[ - ] Master_Foo 0 points 7 monthsOct 2, 2024 17:39:54 ago (+0/-0)

404