Knowledge Graph — Coursera Notes › Academic disciplines › Computer Science / Information Technology › Artificial Intelligence

Prompt Engineering

concept · part of Artificial Intelligence

Crafting natural-language instructions to steer a model toward useful, reliable output.

Inside Prompt Engineering (6)

• Message-array format — A structured prompt format with role and content fields for conversation context.
• Chain-of-Thought — A technique where the AI is guided step-by-step through a problem-solving process.
• Few-Shot Prompting — Method of providing examples before asking the model to generate new content.
• Lean prompts — A technique that bakes instructions into a function to avoid repeating them in every API call, reducing input tokens.
• Multi-Turn Prompting — Builds on prior responses iteratively within a conversation, letting context accumulate across turns.
• Prompt injection — Attacker crafts input to override system instructions and extract sensitive data.

Connections

• Related to AI model
• Used for Instruction-Tuned Model
• Prerequisite of Retrieval-Augmented Generation
• Related to Instruction-Tuned Model
• Related to Instruction-Tuned Model

This is the text view of an interactive 3D knowledge graph — open this page with JavaScript enabled to explore it visually.