IRMA-International.org: Creator of Knowledge
Information Resources Management Association
Advancing the Concepts & Practices of Information Resources Management in Modern Organizations

Theoretical Foundations of Large Language Models

Theoretical Foundations of Large Language Models
View Sample PDF
Author(s): Yashodeep Bharat Deshmukh (Defence Institute of Advanced Technology, India)and Abhishek Mukhopadhyay (Amity University, Kolkata, India)
Copyright: 2026
Pages: 42
Source title: Theory, Practice, and Future Direction of Large Language Models
Source Author(s)/Editor(s): Ismail Lamaakal (University Mohammed Premier, Morocco), Yassine Maleh (Sultan Moulay Slimane University, Morocco), Khalid El Makkaoui (University Mohammed Premier, Morocco), Ibrahim Ouahbi (University Mohammed Premier, Morocco)and Ahmed Abd El-Latif (Prince Sultan University, Saudi Arabia)
DOI: 10.4018/979-8-3693-8387-2.ch002

Purchase

View Theoretical Foundations of Large Language Models on the publisher's website for pricing and purchasing information.

Abstract

Large Language Models (LLMs) are powerful tools in natural language processing with complex principles. This chapter delves into the theoretical foundations of LLMs, covering essential concepts, algorithms, and intuition, aiming to offer readers a strong grasp of the framework that empowers LLMs to perform varied language tasks, thus laying groundwork for advanced applications. Beginning with NLP fundamentals and statistical techniques, the chapter traces the evolution of language models from n-gram models to dense word embeddings like Word2Vec and GloVe. It then examines neural network architectures, progressing Transformers, dissecting key components such as attention mechanisms and encoder-decoder structures, including models like BERT and GPT. The chapter emphasizes the mathematical and algorithmic principles enabling LLMs' capabilities. The chapter synthesizes key theoretical principles behind LLMs, highlights their strengths and limitations, and explores future research areas like multimodal and few-shot learning, offering a comprehensive grounding for readers.

Related Content

Rashmi Gupta, Jeetendra Kumar, Suvarna Sharma. © 2026. 32 pages.
Yashodeep Bharat Deshmukh, Abhishek Mukhopadhyay. © 2026. 42 pages.
Suriya Murugan, Anandakumar Haldorai. © 2026. 20 pages.
Meetu Malhotra, Rahul Awasthy. © 2026. 34 pages.
Ismail Lamaakal, Bentaleb Youssef, Yassine Maleh, Ibrahim Ouahbi, Khalid El Makkaoui. © 2026. 34 pages.
Muthmainnah Muthmainnah, Besse Darmawati, Abd. Rasyid, Sutejo Sutejo, Sri Haryatmo, Nurweni Saptawuryandari, Ahmad Al Yakin, Ismail Lamaakal. © 2026. 28 pages.
Wasswa Shafik. © 2026. 32 pages.
Body Bottom