# Fortifying Toxic Speech Detectors Against Veiled Toxicity

arXiv: https://arxiv.org/abs/2010.03154 · HTML: https://mlsanity.com/paper/2010.03154/

## Key facts

- Cited by 7 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 7 Oct 2020.

## Built on by (most cited first, 7 of 7)

- [Can Machines Help Us Answering Question 16 in Datasheets, and In Turn Reflecting on Inappropriate Content?](https://mlsanity.com/paper/2202.06675/) (2022)
- [Training Data Influence Analysis and Estimation: A Survey](https://mlsanity.com/paper/2212.04612/) (2022)
- [SWSR: A Chinese Dataset and Lexicon for Online Sexism Detection](https://mlsanity.com/paper/2108.03070/) (2021)
- [First is Better Than Last for Language Data Influence](https://mlsanity.com/paper/2202.11844/) (2022)
- [Methods for Detoxification of Texts for the Russian Language](https://mlsanity.com/paper/2105.09052/) (2021)
- ["Cold, Calculated, and Condescending": How AI Identifies and Explains Ableism Compared to Disabled People](https://mlsanity.com/paper/2410.03448/) (2024)
- [HMS-BERT: Hybrid Multi-Task Self-Training for Multilingual and Multi-Label Cyberbullying Detection](https://mlsanity.com/paper/2603.12920/) (2026)

## Abstract (arXiv)

Modern toxic speech detectors are incompetent in recognizing disguised offensive language, such as adversarial attacks that deliberately avoid known toxic lexicons, or manifestations of implicit bias. Building a large annotated dataset for such veiled toxicity can be very expensive. In this work, we propose a framework aimed at fortifying existing toxic speech detectors without a large labeled corpus of veiled toxicity. Just a handful of probing examples are used to surface orders of magnitude more disguised offenses. We augment the toxic speech detector's training data with these discovered offensive examples, thereby making it more robust to veiled toxicity while preserving its utility in detecting overt toxicity.