There’s only one mod of [email protected]
I commented on their meme about Kamala Harris being just as likely to commit war crimes as Trump with an admittedly snarky, sarcastic reply that basically said “some of us wanted to whatever we could, as little as it might be, instead of watching the world burn. Must feel real morally superior safe behind that keyboard”
They banned me from the community for it.
Kinda funny for a community that bills itself as “free from the influence of .ml”
I’m not really an expert on all the details. So I might be wrong here. I don’t know the percentages of how much is done in pretraining and how much in tuning. But from what I know the neural pathways are established in the pretraining phase. Reportedly that’s also where the model learns about the concepts it internalises… Where it gets its world knowledge. So it seems to me a complicated process like learning about a concept like a feeling, or an experience would get established in pretraining already. RLHF is more about what it does with it. But the lines between RLHF, fine-tuning and pretraining are a bit blurry anyway. If I had to guess, I’d say qualia is more likely to be disposed early on, while there’s a lot of changes happening to the neural pathways, so in the pretraining. I’m basing that on my belief, that it’ll be a complex concept… But ultimately there’s no good way to tell, because we don’t know how it’d look like for AI.
Furthermore I’d had a bit of a look what weird use cases people have for AI. And I read about the community efforts to make them usable for NSFW stuff. These people teach new concepts to AI models after the fact. Like how human anatomy looks underneath the clothes. The physics of those parts of the body. And turns out it’s a major hassle. It might degrade other things. It might just work for something close to what it’s seen, so obviously the AI didn’t understand the new concept properly… These people tend to fail at more general models, obviously it’s hard for AI to learn more than one new concept at a later stage… All these things lead me to believe later stages of training are a bad time for AI to learn entirely new concepts. It seems it requires the groundworks to be there since pretraining. That’s probably why we can fine-tune it to prefer a certain style, like Van Gogh drawings. Or a certain way to speak like in RLHF. But not a complicated concept like anatomy. Because the Van Gogh drawings were there in the pretraining dataset already. And they cleaned the nudes. So I’d assume another complicated concept like qualia also needs to come early on. Or it won’t happen later.
Edit: YT video about emotion in LLMs and current research: https://m.youtube.com/watch?v=j9LoyiUlv9I