# IYKYK: Using language models to decode extremist cryptolects

arXiv: https://arxiv.org/abs/2506.05635 · HTML: https://mlsanity.com/paper/2506.05635/

## Key facts

- Cited by 3 papers in the map (data through 14 Oct 2026).
- Builds on 17 of its 50 references that are in the map.
- First posted on arXiv 5 Jun 2025.

## Builds on (most cited first, 17 of 17)

- [Unsupervised Cross-lingual Representation Learning at Scale](https://mlsanity.com/paper/1911.02116/) (2019)
- [Scaling Language Models: Methods, Analysis & Insights from Training Gopher](https://mlsanity.com/paper/2112.11446/) (2021)
- [Dolma: an Open Corpus of Three Trillion Tokens for Language Model Pretraining Research](https://mlsanity.com/paper/2402.00159/) (2024)
- [HateCheck: Functional Tests for Hate Speech Detection Models](https://mlsanity.com/paper/2012.15606/) (2020)
- [A Pretrainer's Guide to Training Data: Measuring the Effects of Data Age, Domain Coverage, Quality, & Toxicity](https://mlsanity.com/paper/2305.13169/) (2023)
- [Detoxifying Language Models Risks Marginalizing Minority Voices](https://mlsanity.com/paper/2104.06390/) (2021)
- [Massive Choice, Ample Tasks (MaChAmp): A Toolkit for Multi-task Learning in NLP](https://mlsanity.com/paper/2005.14672/) (2020)
- [Handling and Presenting Harmful Text in NLP Research](https://mlsanity.com/paper/2204.14256/) (2022)
- [Language Models for Text Classification: Is In-Context Learning Enough?](https://mlsanity.com/paper/2403.17661/) (2024)
- [From Dogwhistles to Bullhorns: Unveiling Coded Rhetoric with Language Models](https://mlsanity.com/paper/2305.17174/) (2023)
- [Words as Gatekeepers: Measuring Discipline-specific Terms and Meanings in Scholarly Publications](https://mlsanity.com/paper/2212.09676/) (2022)
- [Multilingual Auxiliary Tasks Training: Bridging the Gap between Languages for Zero-Shot Transfer of Hate Speech Detection Models](https://mlsanity.com/paper/2210.13029/) (2022)
- [Identity Construction in a Misogynist Incels Forum](https://mlsanity.com/paper/2306.15745/) (2023)
- [Cloaked Classifiers: Pseudonymization Strategies on Sensitive Classification Tasks](https://mlsanity.com/paper/2406.17875/) (2024)
- [Beyond Dataset Creation: Critical View of Annotation Variation and Bias Probing of a Dataset for Online Radical Content Detection](https://mlsanity.com/paper/2412.11745/) (2024)
- [Jointly modelling the evolution of social structure and language in online communities](https://mlsanity.com/paper/2409.19243/) (2024)
- [Inducing lexicons of in-group language with socio-temporal context](https://mlsanity.com/paper/2409.19257/) (2024)

## Built on by (most cited first, 3 of 3)

- [Triggered: A Statistical Analysis of Environmental Influences on Extremist Groups](https://mlsanity.com/paper/2602.09289/) (2026)
- [CyrillicQA: The Influence of Phonetically Encoded Secret Language on LLM Performance](https://mlsanity.com/paper/2608.21462/) (2026)
- [SinoGlyphBench: A Diagnostic Benchmark for Chinese Glyph-Level Obfuscation in Language-Model Moderation](https://mlsanity.com/paper/2609.05843/) (2026)

## Abstract (arXiv)

Extremist groups develop complex in-group language, also referred to as cryptolects, to exclude or mislead outsiders. We investigate the ability of current language technologies to detect and interpret the cryptolects of two online extremist platforms. Evaluating eight models across six tasks, our results indicate that general purpose LLMs cannot consistently detect or decode extremist language. However, performance can be significantly improved by domain adaptation and specialised prompting techniques. These results provide important insights to inform the development and deployment of automated moderation technologies. We further develop and release novel labelled and unlabelled datasets, including 19.4M posts from extremist platforms and lexicons validated by human experts.