mlsanityData to 7 Oct 2026

BiasTrace: Linking Reasoning Behaviours to Biased Outputs in LLMs

arXiv 2608.14161 ↗First posted 14 Aug 2026cs.AIcited by 0

On the map: LLM reasoning → Reasoning benchmarks and evaluation

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.

Key facts. No citing papers in the map yet (data through 7 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.

Builds on

19 of 45 references.

Closest ancestors · 6

Same direction, newest first

  1. 2025Is Reasoning All You Need? Probing Bias in the Age of Reasoning Language Models3 Jul
  2. 2025Qwen3 Technical Report14 May
  3. 2025Reasoning Towards Fairness: Mitigating Bias in Language Models through Reasoning-Guided Fine-Tuning8 Apr
  4. 2025Explicit vs. Implicit: Investigating Social Bias in Large Language Models through Self-Reflection4 Jan
  5. 2024Evaluating Gender Bias in Large Language Models via Chain-of-Thought Prompting28 Jan
  6. 2023Bias Runs Deep: Implicit Reasoning Biases in Persona-Assigned LLMs8 Nov

Other references · 13

  1. 2024The Llama 3 Herd of Modelscited by 14,192
  2. 2025DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learningcited by 6,146
  3. 2022Training a Helpful and Harmless Assistant with Reinforcement Learning from Human Feedbackcited by 3,634
  4. 2025gpt-oss-120b & gpt-oss-20b Model Cardcited by 1,258
  5. 2020StereoSet: Measuring stereotypical bias in pretrained language modelscited by 911
  6. 2025DeepSeek-V3.2: Pushing the Frontier of Open Large Language Modelscited by 755
  7. 2021BBQ: A Hand-Built Bias Benchmark for Question Answeringcited by 683
  8. 2025Chain of Thought Monitorability: A New and Fragile Opportunity for AI Safetycited by 222
  9. 2023The Capacity for Moral Self-Correction in Large Language Modelscited by 171
  10. 2025Checklists Are Better Than Reward Models For Aligning Language Modelscited by 115
  11. 2026A Survey of Agentic Reasoning for Large Language Models: Towards Recursively Self-Improving and Collective Agentscited by 47
  12. 2025A Tutorial on LLM Reasoning: Relevant Methods behind ChatGPT o1cited by 24
1 more
  1. 2025Datasets for Fairness in Language Models: An In-Depth Surveycited by 7

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2025Does Reasoning Introduce Bias? A Study of Social Bias Evaluation and Mitigation in LLM Reasoning
  2. 2025Evaluating Implicit Biases in LLM Reasoning through Logic Grid Puzzles
  3. 2026Evaluating Second-Order Bias of LLMs Through Epistemic Entitlement
  4. 2024CBEval: A framework for evaluating and interpreting cognitive biases in LLMs
  5. 2026Adaptive Triggering for Bias Correction in LLM Reasoning
  6. 2026Conditional Cognitive Biases in LLMs: How Biased User Turns Modulate In-Context Reasoning
  7. 2025Veracity Bias and Beyond: Uncovering LLMs' Hidden Beliefs in Problem-Solving Reasoning
  8. 2026How Reasoning Shapes Social Bias in LLM-Generated Code?