Learning from Red Teaming: Gender Bias Provocation and Mitigation in Large Language Models

TL;DR

This paper introduces an automatic test case generation method to detect and mitigate gender bias in large language models, improving fairness without requiring model fine-tuning.

Contribution

It presents the first automated approach for generating bias test cases and demonstrates effective bias mitigation using in-context learning.

Findings

Generated test cases effectively identify gender biases in LLMs.

Using test cases as demonstrations reduces gender bias in model responses.

The approach improves fairness without fine-tuning the models.

Abstract

Recently, researchers have made considerable improvements in dialogue systems with the progress of large language models (LLMs) such as ChatGPT and GPT-4. These LLM-based chatbots encode the potential biases while retaining disparities that can harm humans during interactions. The traditional biases investigation methods often rely on human-written test cases. However, these test cases are usually expensive and limited. In this work, we propose a first-of-its-kind method that automatically generates test cases to detect LLMs' potential gender bias. We apply our method to three well-known LLMs and find that the generated test cases effectively identify the presence of biases. To address the biases identified, we propose a mitigation strategy that uses the generated test cases as demonstrations for in-context learning to circumvent the need for parameter fine-tuning. The experimental results show that LLMs generate fairer responses with the proposed approach.