# Large Language Models Still Exhibit Bias in Long Text

arXiv: https://arxiv.org/abs/2410.17519 · HTML: https://mlsanity.com/paper/2410.17519/

## Key facts

- Cited by 16 papers in the map (data through 14 Oct 2026).
- Builds on 30 of its 39 references that are in the map.
- First posted on arXiv 23 Oct 2024.

## Builds on (most cited first, 30 of 30)

- [GPT-4 Technical Report](https://mlsanity.com/paper/2303.08774/) (2023)
- [The Llama 3 Herd of Models](https://mlsanity.com/paper/2407.21783/) (2024)
- [Llama 2: Open Foundation and Fine-Tuned Chat Models](https://mlsanity.com/paper/2307.09288/) (2023)
- [Judging LLM-as-a-Judge with MT-Bench and Chatbot Arena](https://mlsanity.com/paper/2306.05685/) (2023)
- [Think you have Solved Question Answering? Try ARC, the AI2 Reasoning Challenge](https://mlsanity.com/paper/1803.05457/) (2018)
- [Finetuned Language Models Are Zero-Shot Learners](https://mlsanity.com/paper/2109.01652/) (2021)
- [Sparks of Artificial General Intelligence: Early experiments with GPT-4](https://mlsanity.com/paper/2303.12712/) (2023)
- [G-Eval: NLG Evaluation using GPT-4 with Better Human Alignment](https://mlsanity.com/paper/2303.16634/) (2023)
- [StereoSet: Measuring stereotypical bias in pretrained language models](https://mlsanity.com/paper/2004.09456/) (2020)
- [RACE: Large-scale ReAding Comprehension Dataset From Examinations](https://mlsanity.com/paper/1704.04683/) (2017)
- [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)
- [Gender Bias in Coreference Resolution: Evaluation and Debiasing Methods](https://mlsanity.com/paper/1804.06876/) (2018)
- [Gender Bias in Coreference Resolution](https://mlsanity.com/paper/1804.09301/) (2018)
- [BOLD: Dataset and Metrics for Measuring Biases in Open-Ended Language Generation](https://mlsanity.com/paper/2101.11718/) (2021)
- [On Measuring Social Biases in Sentence Encoders](https://mlsanity.com/paper/1903.10561/) (2019)
- [Measuring Bias in Contextualized Word Representations](https://mlsanity.com/paper/1906.07337/) (2019)
- [Measuring and Reducing Gendered Correlations in Pre-trained Models](https://mlsanity.com/paper/2010.06032/) (2020)
- [Do-Not-Answer: A Dataset for Evaluating Safeguards in LLMs](https://mlsanity.com/paper/2308.13387/) (2023)
- [Process for Adapting Language Models to Society (PALMS) with Values-Targeted Datasets](https://mlsanity.com/paper/2106.10328/) (2021)
- [Detecting Emergent Intersectional Biases: Contextualized Word Embeddings Contain a Distribution of Human-like Biases](https://mlsanity.com/paper/2006.03955/) (2020)
- [RedditBias: A Real-World Resource for Bias Evaluation and Debiasing of Conversational Language Models](https://mlsanity.com/paper/2106.03521/) (2021)
- [UnQovering Stereotyping Biases via Underspecified Questions](https://mlsanity.com/paper/2010.02428/) (2020)
- [Evaluating and Mitigating Discrimination in Language Model Decisions](https://mlsanity.com/paper/2312.03689/) (2023)
- [On Measuring and Mitigating Biased Inferences of Word Embeddings](https://mlsanity.com/paper/1908.09369/) (2019)
- [Mitigating Language-Dependent Ethnic Bias in BERT](https://mlsanity.com/paper/2109.05704/) (2021)
- [Collecting a Large-Scale Gender Bias Dataset for Coreference Resolution and Machine Translation](https://mlsanity.com/paper/2109.03858/) (2021)
- [Improving Conversational Recommendation Systems via Bias Analysis and Language-Model-Enhanced Data Augmentation](https://mlsanity.com/paper/2310.16738/) (2023)
- [Statistical Equity: A Fairness Classification Objective](https://mlsanity.com/paper/2005.07293/) (2020)
- [The Equity Framework: Fairness Beyond Equalized Predictive Outcomes](https://mlsanity.com/paper/2205.01072/) (2022)

## Built on by (most cited first, 16 of 16)

- [Actions Speak Louder than Words: Agent Decisions Reveal Implicit Biases in Language Models](https://mlsanity.com/paper/2501.17420/) (2025)
- [Towards Trustworthy Retrieval Augmented Generation for Large Language Models: A Survey](https://mlsanity.com/paper/2502.06872/) (2025)
- [SAFEPATH: Preventing Harmful Reasoning in Chain-of-Thought via Early Alignment](https://mlsanity.com/paper/2505.14667/) (2025)
- [Social Bias Benchmark for Generation: A Comparison of Generation and QA-Based Evaluations](https://mlsanity.com/paper/2503.06987/) (2025)
- [A2D: Any-Order, Any-Step Safety Alignment for Diffusion Language Models](https://mlsanity.com/paper/2509.23286/) (2025)
- [FAIRE: Assessing Racial and Gender Bias in AI-Driven Resume Evaluations](https://mlsanity.com/paper/2504.01420/) (2025)
- [SEPS: A Separability Measure for Robust Unlearning in LLMs](https://mlsanity.com/paper/2505.14832/) (2025)
- [Preconditioned Test-Time Adaptation for Out-of-Distribution Debiasing in Narrative Generation](https://mlsanity.com/paper/2603.13683/) (2026)
- [DUSK: Do Not Unlearn Shared Knowledge](https://mlsanity.com/paper/2505.15209/) (2025)
- [MALIBU Benchmark: Multi-Agent LLM Implicit Bias Uncovered](https://mlsanity.com/paper/2507.01019/) (2025)
- [Exploring Gender Bias in Large Language Models: An In-depth Dive into the German Language](https://mlsanity.com/paper/2507.16557/) (2025)
- [GenderBench: Evaluation Suite for Gender Biases in LLMs](https://mlsanity.com/paper/2505.12054/) (2025)
- [Redirected, Not Removed: Task-Dependent Stereotyping Reveals the Limits of LLM Alignments](https://mlsanity.com/paper/2604.02669/) (2026)
- [Large Language Models' Complicit Responses to Illicit Instructions across Socio-Legal Contexts](https://mlsanity.com/paper/2511.20736/) (2025)
- [Analysing LLM Persona Generation and Fairness Interpretation in Polarised Geopolitical Contexts](https://mlsanity.com/paper/2603.22837/) (2026)
- [No-Free-Fairness: Fundamental Limits and Trade-offs in Learning Systems](https://mlsanity.com/paper/2606.17810/) (2026)

## Abstract (arXiv)

Existing fairness benchmarks for large language models (LLMs) primarily focus on simple tasks, such as multiple-choice questions, overlooking biases that may arise in more complex scenarios like long-text generation. To address this gap, we introduce the Long Text Fairness Test (LTF-TEST), a framework that evaluates biases in LLMs through essay-style prompts. LTF-TEST covers 14 topics and 10 demographic axes, including gender and race, resulting in 11,948 samples. By assessing both model responses and the reasoning behind them, LTF-TEST uncovers subtle biases that are difficult to detect in simple responses. In our evaluation of five recent LLMs, including GPT-4o and LLaMa3, we identify two key patterns of bias. First, these models frequently favor certain demographic groups in their responses. Second, they show excessive sensitivity toward traditionally disadvantaged groups, often providing overly protective responses while neglecting others. To mitigate these biases, we propose FT-REGARD, a finetuning approach that pairs biased prompts with neutral responses. FT-REGARD reduces gender bias by 34.6% and improves performance by 1.4 percentage points on the BBQ benchmark, offering a promising approach to addressing biases in long-text generation tasks.