🤖 Large Language Models & Autonomous Agents: Sub-Curriculum Index

Welcome to the Large Language Models & Autonomous Agents curriculum. This track covers the full lifecycle of modern Generative AI: autoregressive LLM architectures, in-context prompt engineering, parameter-efficient fine-tuning (LoRA/QLoRA), autonomous ReAct agents, multi-agent state machines with LangGraph, and enterprise safety guardrails.

Every guide in this series strictly follows a two-part learning format:

  • ⚡ Quick Dive: Comparison matrices, prompt patterns, fine-tuning memory math, and guardrail rules.
  • 📖 Extended Guide: Complete Python code implementations, agent loop state graphs, LoRA training configs, and security architectures.

📚 Curriculum Roadmap

# Guide Primary Topics Covered
01 Large Language Models Architecture Autoregressive decoding, KV-Caching, context windows, tokenizer vocabularies, and compute scaling laws.
02 Prompt Engineering & In-Context Learning Zero-shot, Few-shot, Chain-of-Thought (CoT), Tree-of-Thoughts (ToT), Directional stimulus, and structured JSON output generation.
03 Fine-Tuning: PEFT, LoRA & QLoRA Parameter-Efficient Fine-Tuning, Low-Rank Adaptation (LoRA), 4-bit NF4 Quantization (QLoRA), and Hugging Face peft/trl.
04 AI Agents: ReAct, Planning & Tools Reasoning and Acting (ReAct) execution loops, function calling schemas, tool execution sandboxes, and long-term memory.
05 Multi-Agent Systems & LangGraph Hierarchical supervisor patterns, cyclical state graphs with LangGraph, agent debate, and Human-in-the-Loop workflows.
06 LLM Safety & Guardrails Direct/indirect prompt injection defenses, input/output guardrails (NeMo, Llama Guard), and PII exfiltration mitigation.
⌂ All Guides