Biases in Large Language Models: Origins, Inventory, and Discussion

By Roberto Navigli, Simone Conia, Björn Ross. ACM Journal of Data and Information Quality, 2023

526 citations recorded by the source corpus.

In this article, we introduce and discuss the pervasive issue of bias in the large language models that are currently at the core of mainstream approaches to Natural Language Processing (NLP). We first introduce data selection bias, that is, the bias caused by the choice of texts that make up a training corpus. Then, we survey the different types of social bias evidenced in the text generated by language models trained on such corpora, ranging from gender to age, from sexual orientation to ethnicity, and from religion to culture. We conclude with directions focused on measuring, reducing, and tackling the aforementioned types of bias.

Subfields: Artificial Intelligence

View Paper: https://www.semanticscholar.org/paper/6d0656d9bb60a2bea50c4b894fbcc5d1e32134e7

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

1. Qingyun Wu
2. Chi Wang
3. Erkang Zhu
4. Gagan Bansal
5. Jieyu Zhang

Citations in the corpus

• AutoGen: Enabling Next-Gen LLM Applications via Multi-Agent Conversation Framework (2023, background): “• Bias and Fairness: LLMs have been shown to exhibit biases present in their training data [35].”
• AutoGen: Enabling Next-Gen LLM Applications via Multi-Agent Conversation (2023, unknown): “• Bias and Fairness: LLMs have been shown to exhibit biases present in their training data (Navigli et al., 2023).”
• Siren's Song in the AI Ocean: A Survey on Hallucination in Large Language Models (2023, unknown): “Such sources may inherently mirror societal biases, gender bias, stereotypes, or discriminatory beliefs (Navigli et al., 2023).”
• On LLMs-Driven Synthetic Data Generation, Curation, and Evaluation: A Survey (2024, unknown): “Recent research has found that LLMs’ knowledge is long-tailed and biased (Navigli et al., 2023; Fei et al., 2023).”
• Multimodal Large Language Models in Health Care: Applications, Challenges, and Future Outlook (2024, unknown): “Mitigating bias and improving fairness within medical M-LLMs necessitates a multifaceted approach centered on 3 pillars: data integrity, model refinement, and comprehensive evaluation [181,182].”