Reducing Large Language Model Bias with Emphasis on 'Restricted Industries': Automated Dataset Augmentation and Prejudice Quantification

2024/03/20 by Devam Mondal, Carlo Lipizzi, Mondal, Devam +1
Computer Science · #Natural Language Processing Techniques

paper · pdf · doi:10.48550/arxiv.2403.13925

Abstract

Despite the growing capabilities of large language models, there exists concerns about the biases they develop. In this paper, we propose a novel, automated mechanism for debiasing through specified dataset augmentation in the lens of bias producers and in the context of 'restricted industries' with limited data. We additionally create two new additional metrics, the mb-index and db-index, to quantify bias, considering the idea that bias occurs due to both intrinsic model architecture and dataset.

Related

• Towards Resource Efficient and Interpretable Bias Mitigation in Large Language Models
• Bias and Fairness in Large Language Models: A Survey
• Bias and Fairness in Large Language Models: A Survey
• Promoting Equality in Large Language Models: Identifying and Mitigating the Implicit Bias based on Bayesian Theory
• GPTBIAS: A Comprehensive Framework for Evaluating Bias in Large Language Models