PrExMe! Large Scale Prompt Exploration of Open Source LLMs for Machine Translation and Summarization Evaluation

Christoph Leiter, Steffen Eger


Abstract
Large language models (LLMs) have revolutionized NLP research. Notably, in-context learning enables their use as evaluation metrics for natural language generation, making them particularly advantageous in low-resource scenarios and time-restricted applications. In this work, we introduce PrExMe, a large-scale Prompt Exploration for Metrics, where we evaluate more than 720 prompt templates for open-source LLM-based metrics on machine translation (MT) and summarization datasets, totalling over 6.6M evaluations. This extensive comparison (1) benchmarks recent open-source LLMs as metrics and (2) explores the stability and variability of different prompting strategies. We discover that, on the one hand, there are scenarios for which prompts are stable. For instance, some LLMs show idiosyncratic preferences and favor to grade generated texts with textual labels while others prefer to return numeric scores. On the other hand, the stability of prompts and model rankings can be susceptible to seemingly innocuous changes. For example, changing the requested output format from “0 to 100” to "-1 to +1” can strongly affect the rankings in our evaluation. Our study contributes to understanding the impact of different prompting approaches on LLM-based metrics for MT and summarization evaluation, highlighting the most stable prompting patterns and potential limitations.
Anthology ID:
2024.emnlp-main.641
Volume:
Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing
Month:
November
Year:
2024
Address:
Miami, Florida, USA
Editors:
Yaser Al-Onaizan, Mohit Bansal, Yun-Nung Chen
Venue:
EMNLP
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
11481–11506
Language:
URL:
https://aclanthology.org/2024.emnlp-main.641
DOI:
10.18653/v1/2024.emnlp-main.641
Bibkey:
Cite (ACL):
Christoph Leiter and Steffen Eger. 2024. PrExMe! Large Scale Prompt Exploration of Open Source LLMs for Machine Translation and Summarization Evaluation. In Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing, pages 11481–11506, Miami, Florida, USA. Association for Computational Linguistics.
Cite (Informal):
PrExMe! Large Scale Prompt Exploration of Open Source LLMs for Machine Translation and Summarization Evaluation (Leiter & Eger, EMNLP 2024)
Copy Citation:
PDF:
https://aclanthology.org/2024.emnlp-main.641.pdf
Software:
 2024.emnlp-main.641.software.zip
Data:
 2024.emnlp-main.641.data.zip