paper-with-me

홈 › Papers

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

2025-02-21 · Xuyang Wu, Jinming Nian, Ting-Ruen Wei, Zhiqiang Tao, Hsin-Tai Wu, Yi Fang

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, using the BBQ dataset to analyze both prediction accuracy and bias. Our study spans a wide range of mainstream reasoning 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 a stereotype-free baseline in most cases, mitigating bias and improving the accuracy of LLM outputs. Code will be released upon paper acceptance.

📄 PDF Abstract BibTeX arXiv:2502.15361

Code (0)

등록된 구현이 없습니다.

Tasks

Math

Similar Papers 제목 키워드 기반

A Japanese Benchmark for Evaluating Social Bias in Reasoning Based on Attribution Theory

2026-04-01 · Taihei Shiotani, Masahiro Kaneko, Naoaki Okazaki arxiv

In enhancing the fairness of Large Language Models (LLMs), evaluating social biases rooted in the cultural contexts of specific linguistic regions is essential. However, most existing Japanese benchmarks heavily rely on …

Social Bias Frames: Reasoning about Social and Power Implications of Language

2019-11-10 · ACL 2020 6 · Maarten Sap, Saadia Gabriel, Lianhui Qin, Dan Jurafsky 외

Warning: this paper contains content that may be offensive or upsetting. Language has the power to reinforce stereotypes and project social biases onto others. At the core of the challenge is that it is rarely what is st…

Debiasing Large Language Models toward Social Factors in Online Behavior Analytics through Prompt Knowledge Tuning

2026-03-28 · Hossein Salemi, Jitin Krishnan, Hemant Purohit arxiv

Attribution theory explains how individuals interpret and attribute others' behavior in a social context by employing personal (dispositional) and impersonal (situational) causality. Large Language Models (LLMs), trained…

Intent Detection

Rethinking LLM Bias Probing Using Lessons from the Social Sciences

2025-02-28 · Kirsten N. Morehouse, Siddharth Swaroop, Weiwei Pan

The proliferation of LLM bias probes introduces three significant challenges: (1) we lack principled criteria for choosing appropriate probes, (2) we lack a system for reconciling conflicting results across probes, and (…

Evaluating Gender Bias in Large Language Models via Chain-of-Thought Prompting

2024-01-28 · Masahiro Kaneko, Danushka Bollegala, Naoaki Okazaki, Timothy Baldwin

There exist both scalable tasks, like reading comprehension and fact-checking, where model performance improves with model size, and unscalable tasks, like arithmetic reasoning and symbolic reasoning, where model perform…

Arithmetic ReasoningFact CheckingReading Comprehension