LLMs are better at this and it’ll probably marginally improve the output quality but there can potentially be hallucinations (false positives or negatives) even

LLMs are better at this and it’ll probably marginally improve the output quality but there can potentially be hallucinations (false positives or negatives) even in this evaluation task.