Introduction to Language Modeling: From N-grams to Neural Networks

Learn about language modeling fundamentals in this comprehensive lecture covering key concepts from padding techniques to advanced neural language models. Begin with a review of assignments before diving into padding methodologies and exploring the limitations of static embeddings, random initialization, and bag-of-words approaches. Trace the evolution of transformer models through to RLHF (Reinforcement Learning from Human Feedback), followed by an in-depth examination of N-gram language models. Conclude with an exploration of neural language models and their applications in modern natural language processing.