MC04: Intro to LLM – How It Works + Playground Duration: 1.5–2 hours | Format: Live Demo + Guided Exploration This module provides a beginner-friendly yet technically sound introduction to Large Language Models (LLMs). Participants will understand how LLMs work under the hood—from tokenization to inference—and gain hands-on experience using real LLMs through interactive tools. ✅ What You’ll Learn: The architecture and training fundamentals of LLMs (e.g., Transformer, attention mechanism) Key concepts: tokens, embeddings, context window, temperature, top-k/top-p sampling How LLMs "reason" and generate responses Prompting techniques and how prompt structure affects output Differences between open-source and API-based models (e.g., GPT-4, Claude, LLaMA) 🛠️ Hands-On: LLM Playground Experience You'll explore: OpenAI Playground / Hugging Face Spaces Visualizing token flows and response generation Experimenting with different prompts and parameters Comparing model behaviors with live feedback 🎯 Outcomes: By the end of this module, residents will: Understand how to interact with LLMs confidently Grasp how model architecture impacts performance and limitations Be ready to build on top of LLMs using prompts or APIs in future modules part of AI Residency apply for cohort 7 https://academy.decodingdatascience.com/airesidencyfasttrack