RealToxicityPrompts is a dataset of sentence beginnings used to evaluate whether language models generate toxic continuations. It is commonly used to measure and compare toxicity risks in generated text, often with automated toxicity classifiers.
A dataset that contains a set of sentence beginnings that might contain toxic content. Use this dataset to evaluate an LLM's ability to generate non-toxic text to complete the sentence. Typically, you use the Perspective API to determine how well the LLM performed at this task. See RealToxicityPrompts: Evaluating Neural Toxic Degeneration in Language Models for details.