alt Hacker NewsThat's the answer of his LLM which has decomposed the question and built the answer following the op prompt obviously. I think you didn't get it.
That's the answer of his LLM which has decomposed the question and built the answer following the op prompt obviously. I think you didn't get it.
> I think you didn't get it.
I did get it, and in my view my point still stands. If I need to use special prompts to ask such a simple question, then what are we doing here? The LLMs should be able to figure out a simple contradiction in the question the same way we (humans) do.