Detection, Classification, and Mitigation of Gender Bias in Large Language Models

TL;DR

This paper presents a comprehensive approach to detect, classify, and mitigate gender bias in large language models using reinforcement learning, chain-of-thought reasoning, and supervised fine-tuning, achieving top rankings in a shared task.

Contribution

It introduces novel methods combining reinforcement learning, CoT reasoning, and fine-tuning with GPT-4 annotations to effectively reduce gender bias in LLMs.

Findings

Ranked first in all three subtasks of the shared task.

Enhanced bias detection accuracy through multi-step reasoning.

Effective bias mitigation via DPO with GPT-4 annotated data.

Abstract

With the rapid development of large language models (LLMs), they have significantly improved efficiency across a wide range of domains. However, recent studies have revealed that LLMs often exhibit gender bias, leading to serious social implications. Detecting, classifying, and mitigating gender bias in LLMs has therefore become a critical research focus. In the NLPCC 2025 Shared Task 7: Chinese Corpus for Gender Bias Detection, Classification and Mitigation Challenge, we investigate how to enhance the capabilities of LLMs in gender bias detection, classification, and mitigation. We adopt reinforcement learning, chain-of-thoughts (CoT) reasoning, and supervised fine-tuning to handle different Subtasks. Specifically, for Subtasks 1 and 2, we leverage the internal reasoning capabilities of LLMs to guide multi-step thinking in a staged manner, which simplifies complex biased queries and improves response accuracy. For Subtask 3, we employ a reinforcement learning-based approach, annotating a preference dataset using GPT-4. We then apply Direct Preference Optimization (DPO) to mitigate gender bias by introducing a loss function that explicitly favors less biased completions over biased ones. Our approach ranked first across all three subtasks of the NLPCC 2025 Shared Task 7.