大语言模型中的解释公平性：对LLMs在不同人口群体中解释决策差异的实证分析
Explanation Fairness in Large Language Models: An Empirical Analysis of Disparities in How LLMs Justify Decisions Across Demographic Groups
AI总结：
本文提出解释公平性分类法，分析大语言模型在不同群体中解释决策的差异，发现模型选择与差异幅度密切相关，提出两种提示方法减少解释差异。
摘要：
大语言模型（LLMs）不仅用于做决定，还用于解释这些决定。尽管人工智能决策公平性已被广泛研究，但人工智能解释的公平性（LLMs是否在不同人口群体中以同等质量、深度、语气和语言复杂性来解释决策）却很少受到关注。本文引入了解释公平性分类法（EFT），该框架包含五个正式定义、可操作的维度：冗余差异、情感差异、知识谦虚差异、决策相关解释差异和词汇复杂性差异。该分类法在80个提示模板、四个后果性决策领域（招聘、医疗分诊、信用评估、法律判断）和五个LLM上进行了受控的实证研究：GPT-4.1、Claude Sonnet、LLaMA 3.3 70B、GPT-OSS 120B和Qwen3 32B。引入了两种新的黑盒指标：谦虚密度分数（HDS）和解释忠实度代理（EFP），一个启发式指标用于衡量决策相关解释的差异。在多达400对提示中，所有八个EFT指标都显示出统计学上显著的差异（Cohen's d范围从小到大，所有p_BH < 10^(-62)）。模型选择与差异幅度密切相关：Qwen3 32B的冗余差异比LLaMA 3.3 70B大5.9倍。两种基于提示的缓解方法显示出对EFP差异的显著减少（78-95%），但对风格维度没有显著影响，这与假设一致，即风格解释不平等在预训练分布中编码，并不能通过部署级别的指令单独解决。提供了一个可重复的测量框架用于解释层面的公平性审计，对人工智能监管和部署实践有影响。
英文摘要：
Large language models (LLMs) are increasingly deployed not only to make decisions but to explain them. While AI decision fairness has been studied extensively, the fairness of AI explanations (whether LLMs justify decisions with equal quality, depth, tone, and linguistic sophistication across demographic groups) has received little attention. This paper introduces the Explanation Fairness Taxonomy (EFT), a framework comprising five formally defined, operationalizable dimensions: Verbosity Disparity, Sentiment Disparity, Epistemic Hedging Disparity, Decision-Linked Explanation Disparity, and Lexical Complexity Disparity. The taxonomy is instantiated in a controlled empirical study across 80 prompt templates, four consequential decision domains (hiring, medical triage, credit assessment, legal judgment), and five LLMs: GPT-4.1, Claude Sonnet, LLaMA 3.3 70B, GPT-OSS 120B, and Qwen3 32B. Two novel black-box metrics are introduced: the Hedging Density Score (HDS) and the Explanation Faithfulness Proxy (EFP), a heuristic indicator of decision-linked explanation variation. Across up to 400 prompt pairs, all eight EFT metrics show statistically significant disparities (Cohen's d ranging from small to large, all p_BH < 10^(-62)). Model choice is strongly associated with disparity magnitude: Qwen3 32B exhibits verbosity disparities 5.9x larger than LLaMA 3.3 70B. Two prompting-based mitigations show significant reductions in EFP disparity (78-95%) but no significant effect on stylistic dimensions, consistent with the hypothesis that stylistic explanation inequalities are encoded in pre-training distributions and are not resolvable through deployment-level instruction alone. A reproducible measurement framework is offered for explanation-level fairness auditing, with implications for AI regulation and deployment practice.
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。
