Papers
arxiv:2203.09509

ToxiGen: A Large-Scale Machine-Generated Dataset for Adversarial and Implicit Hate Speech Detection

Published on Mar 17, 2022
Authors:
,
,

Abstract

ToxiGen is a large-scale dataset of toxic and benign statements about minority groups, generated by a machine learning model, helping to improve the detection of implicit toxicity and machine-generated hate speech.

Toxic language detection systems often falsely flag text that contains minority group mentions as toxic, as those groups are often the targets of online hate. Such over-reliance on spurious correlations also causes systems to struggle with detecting implicitly toxic language. To help mitigate these issues, we create ToxiGen, a new large-scale and machine-generated dataset of 274k toxic and benign statements about 13 minority groups. We develop a demonstration-based prompting framework and an adversarial classifier-in-the-loop decoding method to generate subtly toxic and benign text with a massive pretrained language model. Controlling machine generation in this way allows ToxiGen to cover implicitly toxic text at a larger scale, and about more demographic groups, than previous resources of human-written text. We conduct a human evaluation on a challenging subset of ToxiGen and find that annotators struggle to distinguish machine-generated text from human-written language. We also find that 94.5% of toxic examples are labeled as hate speech by human annotators. Using three publicly-available datasets, we show that finetuning a toxicity classifier on our data improves its performance on human-written data substantially. We also demonstrate that ToxiGen can be used to fight machine-generated toxicity as finetuning improves the classifier significantly on our evaluation subset. Our code and data can be found at https://github.com/microsoft/ToxiGen.

Community

Sign up or log in to comment

Get this paper in your agent:

hf papers read 2203.09509
Don't have the latest CLI?
curl -LsSf https://hf.co/cli/install.sh | bash

Models citing this paper 339

Browse 339 models citing this paper

Datasets citing this paper 11

Browse 11 datasets citing this paper

Spaces citing this paper 3,362

Browse 3,362 spaces citing this paper

Collections including this paper 2