The training process doesn't preserve information needed for the LLM to infer that. It cannot be anything other than nonsense that sounds plausible, which is what they do best.
The training process doesn't preserve information needed for the LLM to infer that. It cannot be anything other than nonsense that sounds plausible, which is what they do best.