English

Likelihood-based Mitigation of Evaluation Bias in Large Language Models

Computation and Language 2025-11-11 v4 Artificial Intelligence

Abstract

Large Language Models (LLMs) are widely used to evaluate natural language generation tasks as automated metrics. However, the likelihood, a measure of LLM's plausibility for a sentence, can vary due to superficial differences in sentences, such as word order and sentence structure. It is therefore possible that there might be a likelihood bias if LLMs are used for evaluation: they might overrate sentences with higher likelihoods while underrating those with lower likelihoods. In this paper, we investigate the presence and impact of likelihood bias in LLM-based evaluators. We also propose a method to mitigate the likelihood bias. Our method utilizes highly biased instances as few-shot examples for in-context learning. Our experiments in evaluating the data-to-text and grammatical error correction tasks reveal that several LLMs we test display a likelihood bias. Furthermore, our proposed method successfully mitigates this bias, also improving evaluation performance (in terms of correlation of models with human scores) significantly.

Keywords

Cite

@article{arxiv.2402.15987,
  title  = {Likelihood-based Mitigation of Evaluation Bias in Large Language Models},
  author = {Masanari Oi and Masahiro Kaneko and Ryuto Koike and Mengsay Loem and Naoaki Okazaki},
  journal= {arXiv preprint arXiv:2402.15987},
  year   = {2025}
}

Comments

5 main pages

R2 v1 2026-06-28T14:59:20.480Z