Advanced Prompt Engineering: Chain-of-Thought, Few-Shot, and Beyond

Descrizione del corso

A research-grounded course on advanced prompt engineering techniques. Covers chain-of-thought (CoT) and zero-shot CoT, tree-of-thoughts (ToT), self-consistency sampling, least-to-most prompting, generated knowledge prompting, automatic prompt optimization (APO), DSPy-style declarative prompting, constitutional AI and RLHF alignment prompting, adversarial prompt testing, and building systematic prompt evaluation pipelines.

Cosa imparerai

• Curated from public resources. Go beyond basic prompting: CoT, ToT, self-consistency, constitutional AI, and systematic prompt optimization.

Programma del corso

1. Chain-of-Thought prompting: step-by-step reasoning (28 min)
2. Zero-shot CoT: 'Let's think step by step' and variants (22 min)
3. Tree-of-Thoughts: exploring multiple reasoning paths (32 min)
4. Self-consistency: majority voting over multiple samples (20 min)
5. Automatic Prompt Optimization (APO) with DSPy (40 min)
6. Few-shot selection: semantic similarity vs random (25 min)
7. Constitutional AI: value alignment via prompting (35 min)
8. Adversarial prompt testing: red-teaming your own prompts (30 min)

Struttura del corso

2 sezioni, 8 lezioni, 10h 0m di durata complessiva. Il corso si segue interamente online, al proprio ritmo.

Domande frequenti