Chinese Toxic Language Mitigation via Sentiment Polarity Consistent Rewrites

TL;DR

This paper introduces ToxiRewriteCN, a Chinese detoxification dataset, and evaluates 17 language models on their ability to rewrite toxic content while preserving sentiment and intent, highlighting challenges in subtle toxicity cases.

Contribution

The paper presents the first Chinese detoxification dataset with sentiment preservation and provides a comprehensive evaluation of multiple LLMs on detoxification tasks.

Findings

Commercial and MoE models perform best overall.

All models struggle with emoji, homophone, and dialogue-based toxicity.

Balancing safety and emotional fidelity remains challenging.

Abstract

Detoxifying offensive language while preserving the speaker's original intent is a challenging yet critical goal for improving the quality of online interactions. Although large language models (LLMs) show promise in rewriting toxic content, they often default to overly polite rewrites, distorting the emotional tone and communicative intent. This problem is especially acute in Chinese, where toxicity often arises implicitly through emojis, homophones, or discourse context. We present ToxiRewriteCN, the first Chinese detoxification dataset explicitly designed to preserve sentiment polarity. The dataset comprises 1,556 carefully annotated triplets, each containing a toxic sentence, a sentiment-aligned non-toxic rewrite, and labeled toxic spans. It covers five real-world scenarios: standard expressions, emoji-induced and homophonic toxicity, as well as single-turn and multi-turn dialogues. We evaluate 17 LLMs, including commercial and open-source models with variant architectures, across four dimensions: detoxification accuracy, fluency, content preservation, and sentiment polarity. Results show that while commercial and MoE models perform best overall, all models struggle to balance safety with emotional fidelity in more subtle or context-heavy settings such as emoji, homophone, and dialogue-based inputs. We release ToxiRewriteCN to support future research on controllable, sentiment-aware detoxification for Chinese.