The following prompt tests an LLM's ability to perform evaluation on the outputs of two different models as if it was a teacher.
English original
Can you compare the two outputs below as if you were a teacher?
Output from ChatGPT: {output 1}
Output from GPT-4: {output 2}