Large Language Models in Finance: A Survey

What is it about?

This work is a comprehensive survey on the application of Large Language Models (LLMs) in the field of finance. It explores two main aspects: the existing solutions that utilize LLMs for financial tasks and guidance for their adoption in financial applications. The survey provides an overview of how current approaches employ LLMs in finance, including leveraging pre-trained models through zero-shot or few-shot learning, fine-tuning on domain-specific data, and training custom LLMs from scratch. It also evaluates the performance improvements of these models in financial natural language processing tasks. Additionally, the paper proposes a decision framework to assist financial professionals in selecting the appropriate LLM solution based on use case constraints, such as data, compute, and performance needs. This framework offers a pathway from lightweight experimentation to significant investment in customized LLMs. The survey also discusses the limitations and challenges of leveraging LLMs in financial applications and aims to provide a roadmap for responsibly applying LLMs to advance financial AI. This comprehensive survey targets financial professionals, researchers, and developers interested in the intersection of AI and finance.

Featured Image