Capturing Bias Diversity in LLMs

TL;DR

This paper introduces BiasGPT, a framework of multiple customized GPT models reflecting diverse demographic biases, aiming to create more inclusive and representative AI dialogue by merging varied perspectives.

Contribution

The study develops a novel multi-model GPT framework that captures bias diversity, enhancing inclusivity in AI-generated responses.

Findings

Demonstrated GPT models can embed specific demographic biases.

Combined biases produce more inclusive and representative outputs.

Framework enables nuanced AI dialogues reflecting diverse human experiences.

Abstract

This paper presents research on enhancements to Large Language Models (LLMs) through the addition of diversity in its generated outputs. Our study introduces a configuration of multiple LLMs which demonstrates the diversities capable with a single LLM. By developing multiple customised instances of a GPT model, each reflecting biases in specific demographic characteristics including gender, age, and race, we propose, develop and evaluate a framework for a more nuanced and representative AI dialogue which we call BiasGPT. The customised GPT models will ultimately collaborate, merging their diverse perspectives on a topic into an integrated response that captures a broad spectrum of human experiences and viewpoints. In this paper, through experiments, we demonstrate the capabilities of a GPT model to embed different biases which, when combined, can open the possibilities of more inclusive AI technologies.