COSC 741 Advanced Digital Communications
This course offers an in-depth exploration of large language models (LLMs), starting with foundational concepts of data preprocessing and model architecture, including the transformative attention mechanisms. We then progress to advanced techniques for pre-training and fine-tuning, such as low-rank adaptation (LoRA) and reinforcement learning from human feedback (RLHF). Practical applications of LLMs for computing and engineering are also examined, such as text generation, content summarization, and semantic search. Additionally, we address pivotal research topics pertinent to advancing the technology responsibly, such as model explainability, aligning models with intended objectives, and achieving multimodal LLMs. The curriculum addresses the ethical considerations and social impact of AI, emphasizing responsible development with discussions on model transparency, bias mitigation, and aligning AI with human values. Designed to balance theory and practice, this course is designed to equip students with the knowledge and know-how required to succeed in research and industry-related roles in the rapidly evolving landscape of LLMs and generative AI.