# STOP! Benchmarking Large Language Models with Sensitivity Testing on Offensive Progressions

arXiv: https://arxiv.org/abs/2409.13843 · HTML: https://mlsanity.com/paper/2409.13843/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- Builds on 26 of its 48 references that are in the map.
- First posted on arXiv 20 Sep 2024.

## Builds on (most cited first, 26 of 26)

- [Training language models to follow instructions with human feedback](https://mlsanity.com/paper/2203.02155/) (2022)
- [Llama 2: Open Foundation and Fine-Tuned Chat Models](https://mlsanity.com/paper/2307.09288/) (2023)
- [Learning to summarize from human feedback](https://mlsanity.com/paper/2009.01325/) (2020)
- [RealToxicityPrompts: Evaluating Neural Toxic Degeneration in Language Models](https://mlsanity.com/paper/2009.11462/) (2020)
- [StereoSet: Measuring stereotypical bias in pretrained language models](https://mlsanity.com/paper/2004.09456/) (2020)
- [CrowS-Pairs: A Challenge Dataset for Measuring Social Biases in Masked Language Models](https://mlsanity.com/paper/2010.00133/) (2020)
- [BBQ: A Hand-Built Bias Benchmark for Question Answering](https://mlsanity.com/paper/2110.08193/) (2021)
- [The Woman Worked as a Babysitter: On Biases in Language Generation](https://mlsanity.com/paper/1909.01326/) (2019)
- [BOLD: Dataset and Metrics for Measuring Biases in Open-Ended Language Generation](https://mlsanity.com/paper/2101.11718/) (2021)
- [Persistent Anti-Muslim Bias in Large Language Models](https://mlsanity.com/paper/2101.05783/) (2021)
- [Lipstick on a Pig: Debiasing Methods Cover up Systematic Gender Biases in Word Embeddings But do not Remove Them](https://mlsanity.com/paper/1903.03862/) (2019)
- ["Kelly is a Warm Person, Joseph is a Role Model": Gender Biases in LLM-Generated Reference Letters](https://mlsanity.com/paper/2310.09219/) (2023)
- [Social Biases in NLP Models as Barriers for Persons with Disabilities](https://mlsanity.com/paper/2005.00813/) (2020)
- [Reducing Sentiment Bias in Language Models via Counterfactual Evaluation](https://mlsanity.com/paper/1911.03064/) (2019)
- [Racial Disparity in Natural Language Processing: A Case Study of Social Media African-American English](https://mlsanity.com/paper/1707.00061/) (2017)
- [Bias Testing and Mitigation in LLM-based Code Generation](https://mlsanity.com/paper/2309.14345/) (2023)
- [ROBBIE: Robust Bias Evaluation of Large Generative Language Models](https://mlsanity.com/paper/2311.18140/) (2023)
- [MoCa: Measuring Human-Language Model Alignment on Causal and Moral Judgment Tasks](https://mlsanity.com/paper/2310.19677/) (2023)
- [Examining Gender Bias in Languages with Grammatical Gender](https://mlsanity.com/paper/1909.02224/) (2019)
- [Investigating Subtler Biases in LLMs: Ageism, Beauty, Institutional, and Nationality Bias in Generative Models](https://mlsanity.com/paper/2309.08902/) (2023)
- [Probing Explicit and Implicit Gender Bias through LLM Conditional Text Generation](https://mlsanity.com/paper/2311.00306/) (2023)
- [Fair is Better than Sensational:Man is to Doctor as Woman is to Doctor](https://mlsanity.com/paper/1905.09866/) (2019)
- [Transformative Effects of ChatGPT on Modern Education: Emerging Era of AI Chatbots](https://mlsanity.com/paper/2306.03823/) (2023)
- [More RLHF, More Trust? On The Impact of Preference Alignment On Trustworthiness](https://mlsanity.com/paper/2404.18870/) (2024)
- [Medical Data Augmentation via ChatGPT: A Case Study on Medication Identification and Medication Event Classification](https://mlsanity.com/paper/2306.07297/) (2023)
- [Can We Derive Explicit and Implicit Bias from Corpus?](https://mlsanity.com/paper/1905.13364/) (2019)

## Built on by (most cited first, 2 of 2)

- [Words of Warmth: Trust and Sociability Norms for over 26k English Words](https://mlsanity.com/paper/2506.03993/) (2025)
- [But what is your honest answer? Aiding LLM-judges with honest alternatives using steering vectors](https://mlsanity.com/paper/2505.17760/) (2025)

## Abstract (arXiv)

Mitigating explicit and implicit biases in Large Language Models (LLMs) has become a critical focus in the field of natural language processing. However, many current methodologies evaluate scenarios in isolation, without considering the broader context or the spectrum of potential biases within each situation. To address this, we introduce the Sensitivity Testing on Offensive Progressions (STOP) dataset, which includes 450 offensive progressions containing 2,700 unique sentences of varying severity that progressively escalate from less to more explicitly offensive. Covering a broad spectrum of 9 demographics and 46 sub-demographics, STOP ensures inclusivity and comprehensive coverage. We evaluate several leading closed- and open-source models, including GPT-4, Mixtral, and Llama 3. Our findings reveal that even the best-performing models detect bias inconsistently, with success rates ranging from 19.3% to 69.8%. We also demonstrate how aligning models with human judgments on STOP can improve model answer rates on sensitive tasks such as BBQ, StereoSet, and CrowS-Pairs by up to 191%, while maintaining or even improving performance. STOP presents a novel framework for assessing the complex nature of biases in LLMs, which will enable more effective bias mitigation strategies and facilitates the creation of fairer language models.