# BiasTrace: Linking Reasoning Behaviours to Biased Outputs in LLMs

arXiv: https://arxiv.org/abs/2608.14161 · HTML: https://mlsanity.com/paper/2608.14161/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in Reasoning benchmarks and evaluation, one of 17 directions in LLM reasoning.
- Builds on 19 of its 45 references that are in the map.
- First posted on arXiv 14 Aug 2026.

## Placed in

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

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

- [The Llama 3 Herd of Models](https://mlsanity.com/paper/2407.21783/) (2024)
- [Qwen3 Technical Report](https://mlsanity.com/paper/2505.09388/) (2025)
- [DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning](https://mlsanity.com/paper/2501.12948/) (2025)
- [Training a Helpful and Harmless Assistant with Reinforcement Learning from Human Feedback](https://mlsanity.com/paper/2204.05862/) (2022)
- [gpt-oss-120b & gpt-oss-20b Model Card](https://mlsanity.com/paper/2508.10925/) (2025)
- [StereoSet: Measuring stereotypical bias in pretrained language models](https://mlsanity.com/paper/2004.09456/) (2020)
- [DeepSeek-V3.2: Pushing the Frontier of Open Large Language Models](https://mlsanity.com/paper/2512.02556/) (2025)
- [BBQ: A Hand-Built Bias Benchmark for Question Answering](https://mlsanity.com/paper/2110.08193/) (2021)
- [Chain of Thought Monitorability: A New and Fragile Opportunity for AI Safety](https://mlsanity.com/paper/2507.11473/) (2025)
- [Bias Runs Deep: Implicit Reasoning Biases in Persona-Assigned LLMs](https://mlsanity.com/paper/2311.04892/) (2023)
- [The Capacity for Moral Self-Correction in Large Language Models](https://mlsanity.com/paper/2302.07459/) (2023)
- [Checklists Are Better Than Reward Models For Aligning Language Models](https://mlsanity.com/paper/2507.18624/) (2025)
- [Evaluating Gender Bias in Large Language Models via Chain-of-Thought Prompting](https://mlsanity.com/paper/2401.15585/) (2024)
- [A Survey of Agentic Reasoning for Large Language Models: Towards Recursively Self-Improving and Collective Agents](https://mlsanity.com/paper/2601.12538/) (2026)
- [Explicit vs. Implicit: Investigating Social Bias in Large Language Models through Self-Reflection](https://mlsanity.com/paper/2501.02295/) (2025)
- [A Tutorial on LLM Reasoning: Relevant Methods behind ChatGPT o1](https://mlsanity.com/paper/2502.10867/) (2025)
- [Datasets for Fairness in Language Models: An In-Depth Survey](https://mlsanity.com/paper/2506.23411/) (2025)
- [Reasoning Towards Fairness: Mitigating Bias in Language Models through Reasoning-Guided Fine-Tuning](https://mlsanity.com/paper/2504.05632/) (2025)
- [Is Reasoning All You Need? Probing Bias in the Age of Reasoning Language Models](https://mlsanity.com/paper/2507.02799/) (2025)

## Abstract (arXiv)

LLMs exhibit social biases that can produce inaccurate and discriminatory inferences, posing risks in high-stakes applications. While prior work has made progress in measuring and mitigating bias, it largely focuses on final outputs of models, with limited understanding of the mechanisms that produce biased outcomes. Recent advances in LLM reasoning offers a new lens for investigating bias, yet the link between reasoning and bias remains poorly understood. Existing approaches focus primarily on final answer correctness or explicitly biased language, overlooking different behaviours in reasoning that can drive biased outcomes. We introduce BiasTrace, an annotation scheme for labelling reasoning behaviours in model-generated traces and linking them to biased outcomes. BiasTrace captures bias-specific behaviours (e.g., unsupported demographic assumptions) as well as general reasoning patterns that may implicitly contribute to bias (e.g. overthinking). We apply BiasTrace to reasoning traces in bias-sensitive contexts, scaled using validated LLM-as-a-judge methods, producing a large annotated dataset. Our analysis shows that biased outputs often stem from subtle reasoning behaviours rather than explicitly biased language, and that reasoning-level annotations improve bias detection. We further show that BiasTrace behaviours can be exploited for inference-time mitigation. These findings underscore the importance of examining a broader range of reasoning patterns to better understand bias in LLMs.