Why can’t LLM’s like ChatGPT calculate a confidence score when providing an answer to your question and simply reply “I don’t know” instead of hallucinating an answer?

Question

It seems like they all happily make up a completely incorrect answer and never simply say “I don’t know”. It seems like hallucinated answers come when there’s not a lot of information to train them on a topic. Why can’t the model recognize the low amount of training data and generate with a confidence score to determine if they’re making stuff up?

EDIT: Many people point out rightly that the LLMs themselves can’t “understand” their own response and therefore cannot determine if their answers are made up. But I guess the question includes the fact that chat services like ChatGPT already have support services like the Moderation API that evaluate the content of your query and it’s own responses for content moderation purposes, and intervene when the content violates their terms of use. So couldn’t you have another service that evaluates the LLM response for a confidence score to make this work? Perhaps I should have said “LLM chat services” instead of just LLM, but alas, I did not.

In: Technology

Answer 1

Hallucination isn’t a distinct process. The model is working the same in all situations it’s practically speaking always hallucinating.

We just don’t call the answers hallucinations when we like them. But the LLM didn’t do anything differently to get the wrong answer.

It doesn’t know it’s making the wrong shit up as it’s always just making shit up.