Aligning with Human Judgement: The Role of Pairwise Preference in Large Language Model Evaluators

Yinhong Liu, Han Zhou, Zhijiang Guo, Ehsan Shareghi, Ivan Vulić, Anna Korhonen, Nigel Collier

Research Area: Evaluation

Keywords: LLM evaluator; pairwise comparison; human alignment

TL;DR: We discuss the motivation of pairwise preference, formulate the LLM evaluation as a ranking problem and introduce a search-based ranking method that achieves sota performance.

Abstract:

Large Language Models (LLMs) have demonstrated promising capabilities as automatic evaluators in assessing the quality of generated natural language. However, LLMs still exhibit biases in evaluation and often struggle to generate coherent evaluations that align with human assessments. In this work, we first conduct a systematic study of the misalignment between LLM evaluators and human evaluation, revealing that existing calibration methods aimed at mitigating biases of LLMs are insufficient for effectively aligning LLM evaluators. Inspired by the use of preference data in RLHF, we formulate the evaluation as a ranking problem and introduce Pairwise-preference Search (PairS), an uncertainty-guided search method that employs LLMs to conduct pairwise comparisons locally and efficiently ranks candidate texts globally. PairS achieves state-of-the-art performance on representative evaluation tasks in long-form generations and demonstrates significant improvements over direct scoring. Furthermore, we provide insights into the role of pairwise preference in quantifying the
transitivity of LLMs and demonstrate how PairS benefits from calibration using debiased pairwise evaluations.

Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the COLM Code of Ethics on https://colmweb.org/CoE.html

Author Guide: I certify that this submission complies with the submission instructions as described on https://colmweb.org/AuthorGuide.html

Submission Number: 350