# Conditional Reliability of Toxicity Signals for Multilingual and Code-Mixed Abuse Detection

arXiv: https://arxiv.org/abs/2607.15861 · HTML: https://mlsanity.com/paper/2607.15861/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 15 of its 21 references that are in the map.
- First posted on arXiv 17 Jul 2026.

## Builds on (most cited first, 15 of 15)

- [BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding](https://mlsanity.com/paper/1810.04805/) (2018)
- [On Calibration of Modern Neural Networks](https://mlsanity.com/paper/1706.04599/) (2017)
- [Unsupervised Cross-lingual Representation Learning at Scale](https://mlsanity.com/paper/1911.02116/) (2019)
- [How multilingual is Multilingual BERT?](https://mlsanity.com/paper/1906.01502/) (2019)
- [MAD-X: An Adapter-Based Framework for Multi-Task Cross-Lingual Transfer](https://mlsanity.com/paper/2005.00052/) (2020)
- [Nuanced Metrics for Measuring Unintended Bias with Real Data for Text Classification](https://mlsanity.com/paper/1903.04561/) (2019)
- [Gated Multimodal Units for Information Fusion](https://mlsanity.com/paper/1702.01992/) (2017)
- [A New Generation of Perspective API: Efficient Multilingual Character-level Transformers](https://mlsanity.com/paper/2202.11176/) (2022)
- [MuRIL: Multilingual Representations for Indian Languages](https://mlsanity.com/paper/2103.10730/) (2021)
- [Understanding Abuse: A Typology of Abusive Language Detection Subtasks](https://mlsanity.com/paper/1705.09899/) (2017)
- [A Survey of Code-switched Speech and Language Processing](https://mlsanity.com/paper/1904.00784/) (2019)
- [Designing Toxic Content Classification for a Diversity of Perspectives](https://mlsanity.com/paper/2106.04511/) (2021)
- [Efficient Large-Scale Multi-Modal Classification](https://mlsanity.com/paper/1802.02892/) (2018)
- [Data Bootstrapping Approaches to Improve Low Resource Abusive Language Detection for Indic Languages](https://mlsanity.com/paper/2204.12543/) (2022)
- [Explain Thyself Bully: Sentiment Aided Cyberbullying Detection with Explanation](https://mlsanity.com/paper/2401.09023/) (2024)

## Abstract (arXiv)

Moderation systems increasingly rely on external toxicity tools, but those tools are unreliable under code-mixing, transliteration, slang, and language mismatch. We study the conditional reliability of toxicity priors in Indian multilingual and code-mixed short text: English toxicity, Indic abuse, and rule-based severity cues can be useful evidence, but only in some linguistic and abuse-severity contexts. We propose ToxGate, a trust-fusion head that conditions each auxiliary signal on the encoder representation before adding it to the prediction state. Across three short-text abuse datasets, four transformer encoders, and five seeds per setting, ToxGate improves over matched plain encoders in 10 of 12 in-domain settings and 7 of 8 transfer settings. The largest and most interpretable gains occur in high-risk moderation slices, including explicit slurs, violent threats, and cross-dataset transfer. The broader lesson is that moderation systems should treat external toxicity tools and priors as conditional evidence rather than fixed features or ground truth, in focused ablations, source-specific gating gives the strongest results in transfer, severe-abuse slices, and high-risk triage.