# Analyzing Toxicity in Deep Conversations: A Reddit Case Study

arXiv: https://arxiv.org/abs/2404.07879 · HTML: https://mlsanity.com/paper/2404.07879/

## Key facts

- Cited by 3 papers in the map (data through 14 Oct 2026).
- Placed in Safety alignment, refusal and guardrails, one of 12 directions in LLM safety and security.
- Builds on 18 of its 38 references that are in the map.
- First posted on arXiv 11 Apr 2024.

## Placed in

- LLM safety and security / [Safety alignment, refusal and guardrails](https://mlsanity.com/safety/safety-alignment-refusal-and-guardrails/)

## Builds on (most cited first, 18 of 18)

- [RoBERTa: A Robustly Optimized BERT Pretraining Approach](https://mlsanity.com/paper/1907.11692/) (2019)
- [Automated Hate Speech Detection and the Problem of Offensive Language](https://mlsanity.com/paper/1703.04009/) (2017)
- [ToxiGen: A Large-Scale Machine-Generated Dataset for Adversarial and Implicit Hate Speech Detection](https://mlsanity.com/paper/2203.09509/) (2022)
- [Nuanced Metrics for Measuring Unintended Bias with Real Data for Text Classification](https://mlsanity.com/paper/1903.04561/) (2019)
- [Large Scale Crowdsourcing and Characterization of Twitter Abusive Behavior](https://mlsanity.com/paper/1802.00393/) (2018)
- [Learning from the Worst: Dynamically Generated Datasets to Improve Online Hate Detection](https://mlsanity.com/paper/2012.15761/) (2020)
- [Latent Hatred: A Benchmark for Understanding Implicit Hate Speech](https://mlsanity.com/paper/2109.05322/) (2021)
- [HateCheck: Functional Tests for Hate Speech Detection Models](https://mlsanity.com/paper/2012.15606/) (2020)
- [Ex Machina: Personal Attacks Seen at Scale](https://mlsanity.com/paper/1610.08914/) (2016)
- [Racial Bias in Hate Speech and Abusive Language Detection Datasets](https://mlsanity.com/paper/1905.12516/) (2019)
- [RedditBias: A Real-World Resource for Bias Evaluation and Debiasing of Conversational Language Models](https://mlsanity.com/paper/2106.03521/) (2021)
- [Demoting Racial Bias in Hate Speech Detection](https://mlsanity.com/paper/2005.12246/) (2020)
- [Sarcasm Analysis using Conversation Context](https://mlsanity.com/paper/1808.07531/) (2018)
- [CoSyn: Detecting Implicit Hate Speech in Online Conversations Using a Context Synergized Hyperbolic Network](https://mlsanity.com/paper/2303.03387/) (2023)
- [Public Wisdom Matters! Discourse-Aware Hyperbolic Fourier Co-Attention for Social-Text Classification](https://mlsanity.com/paper/2209.13017/) (2022)
- [Successful New-entry Prediction for Multi-Party Online Conversations via Latent Topics and Discourse Modeling](https://mlsanity.com/paper/2108.06975/) (2021)
- [Twitter Users' Behavioral Response to Toxic Replies](https://mlsanity.com/paper/2210.13420/) (2022)
- [Comparing Toxicity Across Social Media Platforms for COVID-19 Discourse](https://mlsanity.com/paper/2302.14270/) (2023)

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

- [The Moltbook Files: A Harmless Slopocalypse or Humanity's Last Experiment](https://mlsanity.com/paper/2605.07462/) (2026)
- [ProvocationProbe: Instigating Hate Speech Dataset from Twitter](https://mlsanity.com/paper/2410.19687/) (2024)
- [Article and Comment Frames Shape the Quality of Online Comments](https://mlsanity.com/paper/2603.27889/) (2026)

## Abstract (arXiv)

Online social media has become increasingly popular in recent years due to its ease of access and ability to connect with others. One of social media's main draws is its anonymity, allowing users to share their thoughts and opinions without fear of judgment or retribution. This anonymity has also made social media prone to harmful content, which requires moderation to ensure responsible and productive use. Several methods using artificial intelligence have been employed to detect harmful content. However, conversation and contextual analysis of hate speech are still understudied. Most promising works only analyze a single text at a time rather than the conversation supporting it. In this work, we employ a tree-based approach to understand how users behave concerning toxicity in public conversation settings. To this end, we collect both the posts and the comment sections of the top 100 posts from 8 Reddit communities that allow profanity, totaling over 1 million responses. We find that toxic comments increase the likelihood of subsequent toxic comments being produced in online conversations. Our analysis also shows that immediate context plays a vital role in shaping a response rather than the original post. We also study the effect of consensual profanity and observe overlapping similarities with non-consensual profanity in terms of user behavior and patterns.