Re:MonadGPT – What would have happened if ChatGPT was invent...
by cubefox on Sunday November 26, 2023 @06:46AM
(#38420905)
If LLMs from an internal world model, there are definitely things they "believe", and things they can pretend to believe.
Re:MonadGPT – What would have happened if ChatGPT was invent...
by marci on Sunday November 26, 2023 @09:33AM
(#38421856)
If LLMs have models (tokens), there are definitely tokens inside said models, and a program can manipulate them.
Re:MonadGPT – What would have happened if ChatGPT was invent...
by thomasahle on Sunday November 26, 2023 @10:48AM
(#38422452)
It's more like when you prompt the model with "be an assistant from the time of Shakespeare", you don't know if it'll imitate actual Shakespearean data in its training, or the sea of modern human imitations or "fan fiction".
There are various ways modern knowledge may have "snuck in" through data contamination. If we really want to know what a Shakespearean chatbot would have looked like, we need to cap the training data.
Maybe in the future, using better mechanistic interpretability we can get around this, but not right now.
It's like, when you use prompts like : "You're an helpful assistant", is it believing, pretending, or beleiving to be pretending to be an helpful assistant?
It's as funny as disconcerting to see intent and will attributed to probabiltities. Feels sometimes like we're close to making a religion out of this. History of the human race, i guess (https://www.youtube.com/watch?v=xuCn8ux2gbs for the ref).