this post was submitted on 02 Oct 2026
43 points (92.2% liked)
Technology
88816 readers
3760 users here now
This is a most excellent place for technology news and articles.
Our Rules
- Follow the lemmy.world rules.
- Only tech related news or articles.
- Be excellent to each other!
- Mod approved content bots can post up to 10 articles per day.
- Threads asking for personal tech support may be deleted.
- Politics threads may be removed.
- No memes allowed as posts, OK to post as comments.
- Only approved bots from the list below, this includes using AI responses and summaries. To ask if your bot can be added please contact a mod.
- Check for duplicates before posting, duplicates may be removed
- Accounts 7 days and younger will have their posts automatically removed.
Approved Bots
founded 3 years ago
MODERATORS
you are viewing a single comment's thread
view the rest of the comments
view the rest of the comments
For training maybe. But execution as far as i know are just a bunch of probabilities chucked into matrices. Back in my physcos days I could u derstand the math, but not today.
Also what needs citing, that nodes translate to debugable, reproducible outputs.
Because again, it's all probablilties under the hood.
I mean probabilities in matrices are nodes in a neural net, right?
The only thing that makes it non-deterministic is the tempature value which is known after the fact from my understanding, so that should be able to deterministic.
No.... No they're not. Either that or nerual nets are even dumber than I thought. My understanding was emulating nodes of information like brains do. Thats not probabilities. But willing to be wrong if you can bring sources. The connection between points in probability doesn't make sense. Thats a nonsense statement.
Thats also no how simulated annealing works. The temperature value is roughly the probability it will pick a different, less optimal step, in order to try and find better alternative paths. The temperature value is roughly the probability of trying something 'random'. Not deterministic at all. The temperature value is lowered with time and progression. Its known the entire time, but its still a weighted coin flip that determines which path to take.
https://en.wikipedia.org/wiki/Simulated_annealing
Neural nets are dumber than you thought for sure. What I mean is the coin flip results are known after the fact.
I'm not convinced of that you do. Nor is knowing that flip helpful. Because it's non deterministic. An internal coin flip is not an input either.
What I mean is for post hoc analysis since you know the coin flips and the inputs you can step through the layers and nodes calculation deterministically.
I mean sure, but to what end?
You can see where it went wrong. Not why, how
You can try to reason about it which is an arguable improvement from the former engineer you can't get ahold of anymore. So you can could see that it triggered a node associated with c++ when it generated this string of tokens or compare it other generations to map it to a type of call.
Where it doesn't compare to a human engineer at the moment is quality or novelty.
But you can't?
It didn't reason its way down a branch, it made coinflips all along the way. So its not actually thinking, just weighted coin flips. There's no 'why' it chose a branch. Just the result of a coin flip.
You can't ask a coin why it chose heads. You can unroll the physics of it all day long til the cow comes home. But it still doesn't help you on the next one.
It's not really coin flips it and random adjustments of weighted values that decides which nodes are activated which, plus random adjustments, influences the next node activation until out
Statistical outputs are not as interesting as you think?
Like sure, but its still just a probabilistic outcome. Not one reasoned at any step.
I mean it is a semantic collection of weights tied to tokens (fractions of written language). There is also tool calls which could be analysied for additional input analys.