Exploring Prompting Large Language Models as Explainable Metrics

Abstract: This paper describes the IUST NLP Lab submission to the Prompting LLMs as Explainable Metrics Shared Task at the Eval4NLP 2023 Workshop on Evaluation & Comparison of NLP Systems. We have proposed a zero-shot prompt-based strategy for explainable evaluation of the summarization task using LLMs. The conducted experiments demonstrate the promising potential of LLMs as evaluation metrics in NLP, particularly in the field of summarization. Both few-shot and zero-shot approaches are employed in these experiments. The performance of our best provided prompts achieved a Kendall correlation of 0.477 with human evaluations in the text summarization task on the test data. Code and results are publicly available on GitHub.