• DOI:10.18653/v1/2020.findings-emnlp.301
Corpus ID: 221878771

RealToxicityPrompts: Evaluating Neural Toxic Degeneration in Language Models

@inproceedings{Gehman2020RealToxicityPromptsEN,
  title={RealToxicityPrompts: Evaluating Neural Toxic Degeneration in Language Models},
  author={Samuel Gehman and Suchin Gururangan and Maarten Sap and Yejin Choi and Noah A. Smith},
  booktitle={EMNLP},
  year={2020}
}

Samuel Gehman, Suchin Gururangan, +2 authors Noah A. Smith

Published in EMNLP 2020

Computer Science

Pretrained neural language models (LMs) are prone to generating racist, sexist, or otherwise toxic language which hinders their safe deployment. We investigate the extent to which pretrained LMs can be prompted to generate toxic language, and the effectiveness of controllable text generation algorithms at preventing such toxic degeneration. We create and release RealToxicityPrompts, a dataset of 100K naturally occurring, sentence-level prompts derived from a large corpus of English web text… CONTINUE READING

View on ACL

aclweb.org

Save to Library

Create Alert

Cite

Launch Research Feed