LLMs并非银弹:软件公平性的一项案例研究
LLMs Are Not a Silver Bullet: A Case Study on Software Fairness
AI总结:
本文通过对比传统机器学习与大语言模型在软件公平性中的表现,发现传统方法在公平性和预测性能上均优于LLM,指出LLM在现实数据中效果有限。
摘要:
公平性是人类相关、高风险软件系统的关键要求,推动了大量关于偏见缓解的研究。以往研究多集中在表格数据上使用传统机器学习方法。随着大语言模型的兴起,近期研究开始探索其在相同场景下的偏见缓解应用。然而,LLM方法是否优于传统机器学习方法仍不明确。为此,本文进行了大规模研究,比较了最先进的机器学习和LLM偏见缓解方法。发现机器学习方法在公平性和预测性能上均优于LLM方法,即使强LLM也未能超越传统机器学习基线。为理解为何先前LLM研究报告良好结果,我们分析了其评估设置,发现这些收益主要由人工平衡测试数据驱动,而非现实不平衡分布。进一步发现现有LLM方法主要依赖上下文学习,无法利用全部训练数据。受此启发,我们探索了全训练集上的监督微调,发现虽然结果具有竞争力,但其优势仍有限于传统机器学习方法。这些发现表明,LLM并非软件公平性的银弹。
英文摘要:
Fairness is a critical requirement for human-related, high-stakes software systems, motivating extensive research on bias mitigation. Prior work has largely focused on tabular data settings using traditional Machine Learning (ML) methods. With the rapid rise of Large Language Models (LLMs), recent studies have begun to explore their use for bias mitigation in the same setting. However, it remains unclear whether LLM-based methods offer advantages over traditional ML methods, leaving software engineers without clear guidance for practical adoption. To address this gap, we present a large-scale study comparing state-of-the-art ML- and LLM-based bias mitigation methods. We find that ML-based methods consistently outperform LLM-based methods in both fairness and predictive performance, with even strong LLMs failing to surpass established ML baselines. To understand why prior LLM-based studies report favorable results, we analyze their evaluation settings and show that these gains are largely driven by artificially balanced test data rather than realistic imbalanced distributions. We further observe that existing LLM-based methods primarily rely on in-context learning and thus fail to leverage all available training data. Motivated by this, we explore supervised fine-tuning on the full training set and find that, while it achieves competitive results, its advantages over traditional ML methods remain limited. These findings suggest that LLMs are not a silver bullet for software fairness.
机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。