# Does Reasoning Introduce Bias? A Study of Social Bias Evaluation and Mitigation in LLM Reasoning

arXiv: https://arxiv.org/abs/2502.15361 · HTML: https://mlsanity.com/paper/2502.15361/

## Key facts

- Cited by 21 papers in the map (data through 14 Oct 2026).
- Placed in Reasoning benchmarks and evaluation, one of 17 directions in LLM reasoning.
- Builds on 13 of its 39 references that are in the map.
- First posted on arXiv 21 Feb 2025.

## Placed in

- LLM reasoning / [Reasoning benchmarks and evaluation](https://mlsanity.com/lr/reasoning-benchmarks-and-evaluation/)

## Builds on (most cited first, 13 of 13)

- [The Llama 3 Herd of Models](https://mlsanity.com/paper/2407.21783/) (2024)
- [DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning](https://mlsanity.com/paper/2501.12948/) (2025)
- [Qwen2 Technical Report](https://mlsanity.com/paper/2407.10671/) (2024)
- [Monte Carlo Tree Search Boosts Reasoning via Iterative Preference Learning](https://mlsanity.com/paper/2405.00451/) (2024)
- [Thoughts Are All Over the Place: On the Underthinking of o1-Like LLMs](https://mlsanity.com/paper/2501.18585/) (2025)
- [Marco-o1: Towards Open Reasoning Models for Open-Ended Solutions](https://mlsanity.com/paper/2411.14405/) (2024)
- [Evaluating Gender Bias in Large Language Models via Chain-of-Thought Prompting](https://mlsanity.com/paper/2401.15585/) (2024)
- [Decoding Biases: Automated Methods and LLM Judges for Gender Bias Detection in Language Models](https://mlsanity.com/paper/2408.03907/) (2024)
- [Prompting Frameworks for Large Language Models: A Survey](https://mlsanity.com/paper/2311.12785/) (2023)
- [Mathematical Language Models: A Survey](https://mlsanity.com/paper/2312.07622/) (2023)
- [Evaluating Fairness in Large Vision-Language Models Across Diverse Demographic Attributes and Prompts](https://mlsanity.com/paper/2406.17974/) (2024)
- [A Survey on Feedback-based Multi-step Reasoning for Large Language Models on Mathematics](https://mlsanity.com/paper/2502.14333/) (2025)
- [Evaluating and Mitigating Social Bias for Large Language Models in Open-ended Settings](https://mlsanity.com/paper/2412.06134/) (2024)

## Built on by (most cited first, 21 of 21)

- [Self-Preference Bias in Rubric-Based Evaluation of Large Language Models](https://mlsanity.com/paper/2604.06996/) (2026)
- [Schoenfeld's Anatomy of Mathematical Reasoning by Language Models](https://mlsanity.com/paper/2512.19995/) (2025)
- [Reward Modeling for Reinforcement Learning-Based LLM Reasoning: Design, Challenges, and Evaluation](https://mlsanity.com/paper/2602.09305/) (2026)
- [Is Reasoning All You Need? Probing Bias in the Age of Reasoning Language Models](https://mlsanity.com/paper/2507.02799/) (2025)
- [Cognitive Chain-of-Thought (CoCoT): Structured Multimodal Reasoning about Social Situations](https://mlsanity.com/paper/2507.20409/) (2025)
- [Test-Time Reasoners Are Strategic Multiple-Choice Test-Takers](https://mlsanity.com/paper/2510.07761/) (2025)
- [BiasCause: Evaluate Socially Biased Causal Reasoning of Large Language Models](https://mlsanity.com/paper/2504.07997/) (2025)
- [Think Again! The Effect of Test-Time Compute on Preferences, Opinions, and Beliefs of Large Language Models](https://mlsanity.com/paper/2505.19621/) (2025)
- [Investigating Thinking Behaviours of Reasoning-Based Language Models for Social Bias Mitigation](https://mlsanity.com/paper/2510.17062/) (2025)
- [Inference-Time Reasoning Selectively Reduces Implicit Social Bias in Large Language Models](https://mlsanity.com/paper/2602.04742/) (2026)
- [Calibrated? Not for Everyone: How Sexual Orientation and Religious Markers Distort LLM Accuracy and Confidence in Medical QA](https://mlsanity.com/paper/2604.17316/) (2026)
- [Taming Actor-Observer Asymmetry in Agents via Dialectical Alignment](https://mlsanity.com/paper/2604.19548/) (2026)
- [Evaluating Implicit Biases in LLM Reasoning through Logic Grid Puzzles](https://mlsanity.com/paper/2511.06160/) (2025)
- [QuantiBias: Benchmarking Quantization-Induced Bias in LLMs](https://mlsanity.com/paper/2607.21063/) (2026)
- [The Periodic Table of LLM Reasoning: A Structured Survey of Reasoning Paradigms, Methods, and Failure Modes](https://mlsanity.com/paper/2606.11470/) (2026)
- [Evaluating Second-Order Bias of LLMs Through Epistemic Entitlement](https://mlsanity.com/paper/2606.17506/) (2026)
- [Wait, am I Being Fair? Characterizing Deductive Stereotyping and Mitigating It with Fair-GCG](https://mlsanity.com/paper/2606.30989/) (2026)
- [Position: Fairness Failure in Generative Models is an Evaluation Problem](https://mlsanity.com/paper/2608.16974/) (2026)
- [Adaptive Triggering for Bias Correction in LLM Reasoning](https://mlsanity.com/paper/2608.25379/) (2026)
- [When Who You Are Can Change the Code You Get: A Study of Persona-Induced Bias in LLM Code Generation](https://mlsanity.com/paper/2609.22102/) (2026)
- [Does Thinking Help Fairness? Reasoning Tokens Resolve Some Biases but Create More](https://mlsanity.com/paper/2609.30768/) (2026)

## Abstract (arXiv)

Recent advances in large language models (LLMs) have enabled automatic generation of chain-of-thought (CoT) reasoning, leading to strong performance on tasks such as math and code. However, when reasoning steps reflect social stereotypes (e.g., those related to gender, race or age), they can reinforce harmful associations and lead to misleading conclusions. We present the first systematic evaluation of social bias within LLM-generated reasoning, focusing on reasoning language models (e.g., DeepSeek-R1, OpenAI o1) that natively produce reasoning chains as part of their answers. Using the BBQ dataset, we analyze both prediction accuracy and reasoning bias across a broad spectrum of models, including instruction-tuned and CoT-augmented variants of DeepSeek-R1 (8B/32B), ChatGPT, and other open-source LLMs. We quantify how biased reasoning steps correlate with incorrect predictions and often lead to stereotype expression. To mitigate reasoning-induced bias, we propose Answer Distribution as Bias Proxy (ADBP), a lightweight mitigation method that detects bias by tracking how model predictions change across incremental reasoning steps. ADBP outperforms Stereotype-free Reasoning Pattern (SfRP) baseline in most cases, mitigating bias and improving the accuracy of LLM outputs. Evaluation and mitigation code is available at https://github.com/elviswxy/LLM_reasoning_bias.