Feb 24
Loading video player...
📌 Video Chapters / Timestamps 00:00 — Introduction 00:04:58 — Generative vs Discriminative Models 00:17:25 — Tokenization 00:26:12 — Vectors & Matrices 00:33:10 — Embeddings & Similarity Search 00:44:42 — LLM Temperature Tuning 00:49:38 — Prompt Engineering 01:00:43 — Context Window 📝 Course Description ---- Large Language Models (LLMs) are transforming AI, NLP, and modern software development — but understanding how they actually work internally is where most beginners get stuck. In LLM Fundamentals – Part 1, we build a strong conceptual foundation by breaking down the core building blocks of LLMs in a clear, beginner-friendly, and research-oriented way. 🚀 What You’ll Learn in This Video 🔹 Tokenization in LLMs ( very Crucial) 1. What tokens are and why LLMs don’t read raw text 2. How words, subwords, and characters become tokens 🔹 Vectors & Multi-Dimensional Embeddings 3. What vectors really mean in AI 4.Why embeddings live in high-dimensional space (128, 768, 1024, etc.) 🔹 Embeddings Explained (With Intuition) 5. How text is converted into numerical meaning 6. Why embeddings capture semantic relationships 🔹 Similarity Search Fundamentals 7. Dot Product similarity 8. Cosine Similarity (with intuition, not just formulas) 9.How AI decides which text is “closer” in meaning 🔹 Temperature in LLMs (hyperparameter tuning) 10. What temperature controls 11. Low vs high temperature behavior 12. Creativity vs determinism in text generation 🔹 Contextual Understanding in LLMs 13. How LLMs understand context 14. Why context matters more than keywords 🔹 Context Window Explained 15.What context window size means 16. Why long conversations get forgotten Practical limitations of LLM memory 🔹 Prompt Engineering Basics 17. How prompts influence LLM behavior 18.How to Write clear, structured, and effective prompts 🔹 Generative vs Discriminative Models 19. Core differences explained simply 20. Why LLMs are generative models 21. Real-world use cases of both **This video is perfect if you are: **A beginner in AI / NLP **A developer working with GPT, LLaMA, Claude, or open-source models ** A CS or ML student, A researcher starting with LLMs