Debiasing word embeddings has been largely limited to individual and
independent social categories. However, real-world corpora typically present
multiple social categories that possibly correlate or intersect with each
other. For instance, "hair weaves" is stereotypically associated with African
American females, but neither African American nor females alone. Therefore,
this work studies biases associated with multiple social categories: joint
biases induced by the union of different categories and intersectional biases
that do not overlap with the biases of the constituent categories. We first
empirically observe that individual biases intersect non-trivially (i.e., over
a one-dimensional subspace). Drawing from the intersectional theory in social
science and the linguistic theory, we then construct an intersectional subspace
to debias for multiple social categories using the nonlinear geometry of
individual biases. Empirical evaluations corroborate the efficacy of our
approach. Data and implementation code can be downloaded at
this https URL.

本文研究了多个社会类别上的偏差问题，并构建了交叉子空间来去除这些交叉偏差，实验证明这种方法的有效性。

使用非线性几何修正词向量偏置

Debiasing Word Embeddings with Nonlinear Geometry

To examine whether intersectional bias can be observed in language
generation, we examine \emph{GPT-2} and \emph{GPT-NEO} models, ranging in size
from 124 million to ~2.7 billion parameters. We conduct an experiment combining
up to three social categories - gender, religion and disability - into
unconditional or zero-shot prompts used to generate sentences that are then
analysed for sentiment. Our results confirm earlier tests conducted with
auto-regressive causal models, including the \emph{GPT} family of models. We
also illustrate why bias may be resistant to techniques that target single
categories (e.g. gender, religion and race), as it can also manifest, in often
subtle ways, in texts prompted by concatenated social categories. To address
these difficulties, we suggest technical and community-based approaches need to
combine to acknowledge and address complex and intersectional language model
bias.

这篇研究论文主要研究了 GPT-2 和 GPT-NEO 语言生成模型中存在的交叉偏见问题，发现单一类别偏见解决方法的局限性，需要采用综合的技术和社区方法来应对复杂的交叉偏见问题。

因果语言模型中的交叉偏差

Intersectional Bias in Causal Language Models

The use of counterfactuals for considerations of algorithmic fairness and
explainability is gaining prominence within the machine learning community and
industry. This paper argues for more caution with the use of counterfactuals
when the facts to be considered are social categories such as race or gender.
We review a broad body of papers from philosophy and social sciences on social
ontology and the semantics of counterfactuals, and we conclude that the
counterfactual approach in machine learning fairness and social explainability
can require an incoherent theory of what social categories are. Our findings
suggest that most often the social categories may not admit counterfactual
manipulation, and hence may not appropriately satisfy the demands for
evaluating the truth or falsity of counterfactuals. This is important because
the widespread use of counterfactuals in machine learning can lead to
misleading results when applied in high-stakes domains. Accordingly, we argue
that even though counterfactuals play an essential part in some causal
inferences, their use for questions of algorithmic fairness and social
explanations can create more problems than they resolve. Our positive result is
a set of tenets about using counterfactuals for fairness and explanations in
machine learning.

本文指出过度使用对社会类别（如种族或性别）进行因果推断所需的反事实可能导致一些困扰，并可能导致错误结论的结果，因此应该谨慎使用因果反事实，特别是在涉及高风险领域的算法公平和社交解释方面。