Certified Advanced Prompting
Multi-step reasoning
Module: Advanced Reasoning & Logic
Break complex problems into ordered reasoning steps the model can verify as it goes.
In practice: 'Solve in steps, checking each before the next.'
Try It Yourself
'Solve in steps, checking each before the next.'
Lessons
▶️ 1. Multi-step reasoning
▶️ 2. Analogical reasoning
▶️ 3. Counterfactual prompting
▶️ 4. Deductive vs inductive
▶️ 5. Constraint satisfaction
▶️ 6. Reasoning with uncertainty
▶️ 7. Error analysis
▶️ 8. Verifying conclusions
▶️ 9. Working with long inputs
▶️ 10. Summarise-then-act
▶️ 11. Rolling summaries
▶️ 12. Reference markers
▶️ 13. Chunking strategy
▶️ 14. External memory
▶️ 15. Avoiding context loss
▶️ 16. Long-context pitfalls
▶️ 17. Strict JSON output
▶️ 18. Function calling
▶️ 19. Schema design
▶️ 20. Validation and retries
▶️ 21. Enums and controlled values
▶️ 22. Parsing model output
▶️ 23. Idempotent prompts
▶️ 24. Integrating with apps
▶️ 25. Diagnosing weak prompts
▶️ 26. Tightening instructions
▶️ 27. Prompt compression
▶️ 28. A/B testing prompts
▶️ 29. Adding constraints to improve quality
▶️ 30. Reducing hallucination
▶️ 31. Token and cost awareness
▶️ 32. Prompt versioning
▶️ 33. Why evaluate prompts
▶️ 34. Building a test set
▶️ 35. Defining quality criteria
▶️ 36. Human evaluation
▶️ 37. Automated evaluation
▶️ 38. LLM-as-judge
▶️ 39. Catching regressions
▶️ 40. Measuring consistency
Course Home