Text Detoxification using Large Pre-trained Neural Models

TL;DR

This paper introduces two innovative unsupervised approaches for removing toxicity from text using large pre-trained neural models, achieving state-of-the-art results in style transfer for toxicity mitigation.

Contribution

The paper proposes two novel unsupervised methods for text detoxification, combining style-guided generation and synonym replacement with BERT, along with the first large-scale comparative evaluation.

Findings

Both methods achieve new state-of-the-art results.

The style-guided paraphrasing effectively reduces toxicity.

BERT-based word replacement improves flexibility and effectiveness.

Abstract

We present two novel unsupervised methods for eliminating toxicity in text. Our first method combines two recent ideas: (1) guidance of the generation process with small style-conditional language models and (2) use of paraphrasing models to perform style transfer. We use a well-performing paraphraser guided by style-trained language models to keep the text content and remove toxicity. Our second method uses BERT to replace toxic words with their non-offensive synonyms. We make the method more flexible by enabling BERT to replace mask tokens with a variable number of words. Finally, we present the first large-scale comparative study of style transfer models on the task of toxicity removal. We compare our models with a number of methods for style transfer. The models are evaluated in a reference-free way using a combination of unsupervised style transfer metrics. Both methods we suggest yield new SOTA results.