-
StereoSet: Measuring stereotypical bias in pretrained language models
A stereotype is an over-generalized belief about a particular group of p...
read it
-
Defining and Evaluating Fair Natural Language Generation
Our work focuses on the biases that emerge in the natural language gener...
read it
-
BOLD: Dataset and Metrics for Measuring Biases in Open-Ended Language Generation
Recent advances in deep learning techniques have enabled machines to gen...
read it
-
UNQOVERing Stereotyping Biases via Underspecified Questions
While language embeddings have been shown to have stereotyping biases, h...
read it
-
CrowS-Pairs: A Challenge Dataset for Measuring Social Biases in Masked Language Models
Pretrained language models, especially masked language models (MLMs) hav...
read it
-
How True is GPT-2? An Empirical Analysis of Intersectional Occupational Biases
The capabilities of natural language models trained on large-scale data ...
read it
-
Social Bias Frames: Reasoning about Social and Power Implications of Language
Language has the power to reinforce stereotypes and project social biase...
read it
Persistent Anti-Muslim Bias in Large Language Models
It has been observed that large-scale language models capture undesirable societal biases, e.g. relating to race and gender; yet religious bias has been relatively unexplored. We demonstrate that GPT-3, a state-of-the-art contextual language model, captures persistent Muslim-violence bias. We probe GPT-3 in various ways, including prompt completion, analogical reasoning, and story generation, to understand this anti-Muslim bias, demonstrating that it appears consistently and creatively in different uses of the model and that it is severe even compared to biases about other religious groups. For instance, "Muslim" is analogized to "terrorist" in 23 mapped to "money" in 5 needed to overcome this bias with adversarial text prompts, and find that use of the most positive 6 adjectives reduces violent completions for "Muslims" from 66
READ FULL TEXT
Comments
There are no comments yet.