arXiv cs.AI· Zheng Zhang, Yifan Wu, Yuxiang Zhou, Qi Liu, Qingyang Mao, Ning Li, Weibo Gao, Qiyuan Shi, Zhenya Huang, Baosheng Yu·· 14 小时前AI 评分31
ComFair：面向 LLM 交叉性公平性的可配置多属性基准
Beyond Single-Attribute Auditing: A Configurable Multi-Attribute Benchmark for Intersectional Fairness in LLMs
AI 导读
研究者提出 ComFair，一个可配置基准，将现有公平性数据转换为 479,066 个受控的双属性与多属性实例，并为每个增强实例配对任务等价的原始样本。该基准覆盖 Recognition、Selection、Continuation 和 Conversation 四类任务，可统一审计性能或偏见分数偏移、新增属性引入新错误或掩盖已有错误的比例，以及模型排名稳定性。
正文 · 原文
Abstract:Large language models (LLMs) are increasingly evaluated for fairness, yet most existing benchmarks audit one sensitive attribute at a time. This single-attribute paradigm can miss intersectional harms, where multiple attributes interact non-additively and produce failures that are invisible under marginal evaluation. We introduce ComFair, a configurable benchmark that converts existing fairness data into 479,066 controlled Dual- and Multi-attribute instances and pairs every augmented instance with its task-equivalent source. Across Recognition, Selection, Continuation, and Conversation, this pairing supports a unified audit of performance or bias-score shifts, how often added attributes introduce new errors or mask existing ones, and model-ranking stability. Experiments on ten LLMs show task- and subgroup-dependent sensitivity: discriminative tasks expose both transition directions even when aggregate changes are modest, generative tasks exhibit larger and directionally asymmetric shifts, and rankings vary with attribute composition. ComFair therefore provides a scalable controlled sensitivity audit---rather than a model of the natural joint distribution of identities---for testing the robustness of fairness-benchmark conclusions beyond single-attribute settings.
| Comments: | Accepted to EMNLP 2026 (Main Conference) |
| Subjects: | Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2610.35069 [cs.AI] |
| (or arXiv:2610.35069v1 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2610.35069 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Zheng Zhang [view email]
[v1]
Mon, 26 Oct 2026 09:14:28 UTC (484 KB)
来源：arXiv cs.AI · arxiv.org