Dissociating language and thought in large language models
Large language models (LLMs) have come closest among all models to date to mastering
human language, yet opinions about their linguistic and cognitive capabilities remain split …
human language, yet opinions about their linguistic and cognitive capabilities remain split …
The gender-binary cycle: The perpetual relations between a biological-essentialist view of gender, gender ideology, and gender-labelling and sorting
T Saguy, M Reifen-Tagar… - … Transactions of the …, 2021 - royalsocietypublishing.org
Gender inequality is one of the most pressing issues of our time. A core factor that feeds
gender inequality is people's gender ideology—a set of beliefs about the proper order of …
gender inequality is people's gender ideology—a set of beliefs about the proper order of …
Gender bias in machine translation
Abstract Machine translation (MT) technology has facilitated our daily tasks by providing
accessible shortcuts for gathering, processing, and communicating information. However, it …
accessible shortcuts for gathering, processing, and communicating information. However, it …
Rethinking multilingual experience through a Systems Framework of Bilingualism
In “The Devil's Dictionary”, Bierce (1911) defined language as “The music with which we
charm the serpents guarding another's treasure.” This satirical definition reflects a core truth …
charm the serpents guarding another's treasure.” This satirical definition reflects a core truth …
Gender stereotypes in natural language: Word embeddings show robust consistency across child and adult language corpora of more than 65 million words
TES Charlesworth, V Yang, TC Mann… - Psychological …, 2021 - journals.sagepub.com
Stereotypes are associations between social groups and semantic attributes that are widely
shared within societies. The spoken and written language of a society affords a unique way …
shared within societies. The spoken and written language of a society affords a unique way …
Evaluating biased attitude associations of language models in an intersectional context
Language models are trained on large-scale corpora that embed implicit biases
documented in psychology. Valence associations (pleasantness/unpleasantness) of social …
documented in psychology. Valence associations (pleasantness/unpleasantness) of social …
Based on billions of words on the internet, people = men
Recent advances have made it possible to precisely measure the extent to which any two
words are used in similar contexts. In turn, this measure of similarity in linguistic context also …
words are used in similar contexts. In turn, this measure of similarity in linguistic context also …
Cultural schemas: What they are, how to find them, and what to do once you've caught one
A Boutyline, LK Soter - American Sociological Review, 2021 - journals.sagepub.com
Cultural schemas are a central cognitive mechanism through which culture affects action. In
this article, we develop a theoretical model of cultural schemas that is better able to support …
this article, we develop a theoretical model of cultural schemas that is better able to support …
Gender bias in masked language models for multiple languages
Masked Language Models (MLMs) pre-trained by predicting masked tokens on large
corpora have been used successfully in natural language processing tasks for a variety of …
corpora have been used successfully in natural language processing tasks for a variety of …
Debiasing isn't enough!--On the Effectiveness of Debiasing MLMs and their Social Biases in Downstream Tasks
M Kaneko, D Bollegala, N Okazaki - arXiv preprint arXiv:2210.02938, 2022 - arxiv.org
We study the relationship between task-agnostic intrinsic and task-specific extrinsic social
bias evaluation measures for Masked Language Models (MLMs), and find that there exists …
bias evaluation measures for Masked Language Models (MLMs), and find that there exists …