The problem with describing factual errors as "hallucinating" is that they're no more or less hallucinations than when it generate correct content. The entire p

The problem with describing factual errors as "hallucinating" is that they're no more or less hallucinations than when it generate correct content. The entire point of these LLMs is that it synthesises content not in the original input. It's always "hallucinating" - its just that sometimes it gets it right and sometimes it gets it wrong.