Generative Large Language Models Explained [AI-eXplained]

Large Language Models (LLMs) such as OpenAI’s ChatGPT have achieved surprisingly huge progresses in the field of Natural Language Processing (NLP). This paper aims to present an immersive introduction to LLMs from the perspective of generative models. The main components of the training process of LLMs are explained, and an example of LLMs for AI-generated contents is given. This short paper is a summary of the interactive full paper online available at IEEE Xplore, in which detailed examples interactively demonstrate the training and working mechanisms of LLMs.

I. Introduction

Recently, generative Large Language Models (LLMs) have made great strides in artificial intelligence, such as ChatGPT.1

[Online]. Available: https://openai.com/blog/chatgpt/

These models, powered by deep learning techniques, have revolutionized various applications, ranging from chatbots and virtual assistants to machine translation and content generation. With their remarkable ability to understand and generate human-like text, LLMs have garnered huge attention not only among researchers and industrial practitioners, but also the general public. The working mechanisms of LLMs are complex and they are often perceived as black-boxes, making it challenging for non-experts to comprehend how they achieve such impressive capabilities.