ReIFE: Re-evaluating Instruction-Following Evaluation
Liu, Yixin, Shi, Kejian, Fabbri, Alexander R., Zhao, Yilun, Wang, Peifeng, Wu, Chien-Sheng, Joty, Shafiq, Cohan, Arman
–arXiv.org Artificial Intelligence
The automatic evaluation of instruction following typically involves using large language models (LLMs) to assess response quality. However, there is a lack of comprehensive evaluation of these LLM-based evaluators across two dimensions: the base LLMs and the evaluation protocols. Therefore, we present a thorough meta-evaluation of instruction following, including 25 base LLMs and 15 recently proposed evaluation protocols, on 4 human-annotated datasets, assessing the evaluation accuracy of the LLM-evaluators. Our evaluation allows us to identify the best-performing base LLMs and evaluation protocols with a high degree of robustness. Moreover, our large-scale evaluation reveals: (1) Base LLM performance ranking remains largely consistent across evaluation protocols, with less capable LLMs showing greater improvement from protocol enhancements; (2) Robust evaluation of evaluation protocols requires many base LLMs with varying capability levels, as protocol effectiveness can depend on the base LLM used; (3) Evaluation results on different datasets are not always consistent, so a rigorous evaluation requires multiple datasets with distinctive features. We release our meta-evaluation suite ReIFE, which provides the codebase and evaluation result collection for more than 500 LLM-evaluator configurations, to support future research in instruction-following evaluation.
arXiv.org Artificial Intelligence
Oct-9-2024
- Genre:
- Research Report > New Finding (1.00)
- Industry:
- Education (0.45)
- Health & Medicine > Therapeutic Area (0.45)
- Technology: