History, Development, and Principles of Large Language Models-An Introductory Survey

  • 2024-09-23 15:12:40
  • Zichong Wang, Zhibo Chu, Thang Viet Doan, Shiwen Ni, Min Yang, Wenbin Zhang
  • 0

Abstract

Language models serve as a cornerstone in natural language processing (NLP),utilizing mathematical methods to generalize language laws and knowledge forprediction and generation. Over extensive research spanning decades, languagemodeling has progressed from initial statistical language models (SLMs) to thecontemporary landscape of large language models (LLMs). Notably, the swiftevolution of LLMs has reached the ability to process, understand, and generatehuman-level text. Nevertheless, despite the significant advantages that LLMsoffer in improving both work and personal lives, the limited understandingamong general practitioners about the background and principles of these modelshampers their full potential. Notably, most LLM reviews focus on specificaspects and utilize specialized language, posing a challenge for practitionerslacking relevant background knowledge. In light of this, this survey aims topresent a comprehensible overview of LLMs to assist a broader audience. Itstrives to facilitate a comprehensive understanding by exploring the historicalbackground of language models and tracing their evolution over time. The surveyfurther investigates the factors influencing the development of LLMs,emphasizing key contributions. Additionally, it concentrates on elucidating theunderlying principles of LLMs, equipping audiences with essential theoreticalknowledge. The survey also highlights the limitations of existing work andpoints out promising future directions.

 

Quick Read (beta)

loading the full paper ...