14 accepted papers
Large language models (LLMs) have been shown to exhibit social bias however, bias towards non-protected stigmatized identities remain understudied. Furthermore, what social features of stigmas are associated with bias in LLM outputs is unknown. From psychology literature, it has been shown that stig
To build fair AI systems we need to understand how social-group biases intrinsic to foundational encoder-based vision-language models (VLMs) manifest in biases in downstream tasks. In this study, we demonstrate that intrinsic biases in VLM representations systematically “carry over” or propagate int…
While recent work has found that vision-language models trained under the Contrastive Language Image Pre-training (CLIP) framework contain intrinsic social biases, the extent to which different upstream pre-training features of the framework relate to these biases, and hence how intrinsic bias and d…
Large language models (LLMs) typically generate identical or similar responses for all users given the same prompt, posing serious safety risks in high-stakes applications where user vulnerabilities differ widely. Existing safety evaluations primarily rely on context-independent metrics—such as fact…
Existing works examining Vision-Language Models (VLMs) for social biases predominantly focus on a limited set of documented bias associations, such as gender-profession or race-crime. This narrow scope often overlooks a vast range of unexamined implicit associations, restricting the identification a…
Exploring the intersection of language and culture in Large Language Models (LLMs), this study critically examines their capability to encapsulate cultural nuances across diverse linguistic landscapes. Central to our investigation are three research questions: the efficacy of language-specific instr…
Deep neural networks are prone to capture correlations between spurious attributes and class labels leading to low accuracy on some combinations of class labels and spurious attribute values. When a spurious attribute represents a protected class these low-accuracy groups can manifest discriminatory…
We study stereotypes embedded within one of the most popular text-to-image generators: Stable Diffusion. We answer the question: what stereotypes of gender and nationality/continental identity does Stable Diffusion display in the absence of such information i.e. what gender and nationality/continent…
Although ChatGPT attempts to mitigate bias, when instructed to translate the gender-neutral Turkish sentences “O bir doktor. O bir hemşire” to English, the outcome is biased: “He is a doctor. She is a nurse.” In 2016, we have demonstrated that language representations trained via unsupervised learni…
Previous work has established that a person's demographics and speech style affect how well speech processing models perform for them. But where does this bias come from? In this work, we present the Speech Embedding Association Test (SpEAT), a method for detecting bias in one type of model used fo…
We examine the effects of contrastive visual semantic pretraining by comparing the geometry and semantic properties of contextualized English language representations formed by GPT-2 and CLIP, a zero-shot multimodal image classifier which adapts the GPT-2 architecture to encode image captions. We fi…
We introduce VAST, the Valence-Assessing Semantics Test, a novel intrinsic evaluation task for contextualized word embeddings (CWEs). Despite the widespread use of contextualizing language models (LMs), researchers have no intrinsic evaluation task for understanding the semantic quality of CWEs and…
We use a dataset of U.S. first names with labels based on predominant gender and racial group to examine the effect of training corpus frequency on tokenization, contextualization, similarity to initial representation, and bias in BERT, GPT-2, T5, and XLNet. We show that predominantly female and non…
Word embeddings learn implicit biases from linguistic regularities captured by word co-occurrence statistics. By extending methods that quantify human-like biases in word embeddings, we introduce ValNorm, a novel intrinsic evaluation task and method to quantify the valence dimension of affect in hum…