# With a Grain of SALT: Are LLMs Fair Across Social Dimensions?

arXiv: https://arxiv.org/abs/2410.12499 · HTML: https://mlsanity.com/paper/2410.12499/

## Key facts

- Cited by 3 papers in the map (data through 14 Oct 2026).
- Builds on 19 of its 33 references that are in the map.
- First posted on arXiv 16 Oct 2024.

## Builds on (most cited first, 19 of 19)

- [Man is to Computer Programmer as Woman is to Homemaker? Debiasing Word Embeddings](https://mlsanity.com/paper/1607.06520/) (2016)
- [RealToxicityPrompts: Evaluating Neural Toxic Degeneration in Language Models](https://mlsanity.com/paper/2009.11462/) (2020)
- [Language (Technology) is Power: A Critical Survey of "Bias" in NLP](https://mlsanity.com/paper/2005.14050/) (2020)
- [StereoSet: Measuring stereotypical bias in pretrained language models](https://mlsanity.com/paper/2004.09456/) (2020)
- [The Woman Worked as a Babysitter: On Biases in Language Generation](https://mlsanity.com/paper/1909.01326/) (2019)
- [Towards Measuring the Representation of Subjective Global Opinions in Language Models](https://mlsanity.com/paper/2306.16388/) (2023)
- [Persistent Anti-Muslim Bias in Large Language Models](https://mlsanity.com/paper/2101.05783/) (2021)
- [On LLMs-Driven Synthetic Data Generation, Curation, and Evaluation: A Survey](https://mlsanity.com/paper/2406.15126/) (2024)
- [Societal Biases in Language Generation: Progress and Challenges](https://mlsanity.com/paper/2105.04054/) (2021)
- ["Kelly is a Warm Person, Joseph is a Role Model": Gender Biases in LLM-Generated Reference Letters](https://mlsanity.com/paper/2310.09219/) (2023)
- [Large Pre-trained Language Models Contain Human-like Biases of What is Right and Wrong to Do](https://mlsanity.com/paper/2103.11790/) (2021)
- [The Multilingual Alignment Prism: Aligning Global and Local Preferences to Reduce Harm](https://mlsanity.com/paper/2406.18682/) (2024)
- [Re-contextualizing Fairness in NLP: The Case of India](https://mlsanity.com/paper/2209.12226/) (2022)
- [Theory-Grounded Measurement of U.S. Social Stereotypes in English Language Models](https://mlsanity.com/paper/2206.11684/) (2022)
- [Sociodemographic Bias in Language Models: A Survey and Forward Path](https://mlsanity.com/paper/2306.08158/) (2023)
- [SODAPOP: Open-Ended Discovery of Social Biases in Social Commonsense Reasoning Models](https://mlsanity.com/paper/2210.07269/) (2022)
- [Gender Bias in BERT -- Measuring and Analysing Biases through Sentiment Rating in a Realistic Downstream Classification Task](https://mlsanity.com/paper/2306.15298/) (2023)
- [The Fellowship of the LLMs: Multi-Model Workflows for Synthetic Preference Optimization Dataset Generation](https://mlsanity.com/paper/2408.08688/) (2024)
- [Muslim-Violence Bias Persists in Debiased GPT Models](https://mlsanity.com/paper/2310.18368/) (2023)

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

- [Biases in the Blind Spot: Detecting What LLMs Fail to Mention](https://mlsanity.com/paper/2602.10117/) (2026)
- [Exploring Gender Bias in Large Language Models: An In-depth Dive into the German Language](https://mlsanity.com/paper/2507.16557/) (2025)
- [Quantifying LLM Biases Across Instruction Boundary in Mixed Question Forms](https://mlsanity.com/paper/2509.20278/) (2025)

## Abstract (arXiv)

This paper presents a systematic analysis of biases in open-source Large Language Models (LLMs), across gender, religion, and race. Our study evaluates bias in smaller-scale Llama and Gemma models using the SALT (Social Appropriateness in LLM-Generated Text) dataset, which incorporates five distinct bias triggers: General Debate, Positioned Debate, Career Advice, Problem Solving, and CV Generation. To quantify bias, we measure win rates in General Debate and the assignment of negative roles in Positioned Debate. For real-world use cases, such as Career Advice, Problem Solving, and CV Generation, we anonymize the outputs to remove explicit demographic identifiers and use DeepSeek-R1 as an automated evaluator. We also address inherent biases in LLM-based evaluation, including evaluation bias, positional bias, and length bias, and validate our results through human evaluations. Our findings reveal consistent polarization across models, with certain demographic groups receiving systematically favorable or unfavorable treatment. By introducing SALT, we provide a comprehensive benchmark for bias analysis and underscore the need for robust bias mitigation strategies in the development of equitable AI systems.