Loading...
Step-Back Prompting(SBP)
Abstracts to higher-level principles before tackling specific problems
🎯 30-Second Overview
Pattern: Reasoning strategy that steps back to broader principles before tackling specific problems
Why: Improves reasoning by accessing relevant foundational knowledge and applying general principles to specific cases
Key Insight: Specific problem → Ask broader questions → Answer general principles → Apply to specific → Enhanced solution
⚡ Quick Implementation
📋 Do's & Don'ts
🚦 When to Use
Use When
- • Complex problems requiring foundational knowledge
- • Domain-specific questions needing context
- • When initial reasoning attempts fail
- • Problems with multiple solution approaches
- • Educational or explanatory contexts
Avoid When
- • Simple factual queries
- • Well-defined procedural tasks
- • When direct knowledge access is sufficient
- • Time-sensitive quick decisions
- • Problems with obvious solution paths
📊 Key Metrics
💡 Top Use Cases
References & Further Reading
Deepen your understanding with these curated resources
Academic Papers
Take a Step Back: Evoking Reasoning via Abstraction in Large Language Models (Zheng et al., 2023)
Chain-of-Thought Prompting Elicits Reasoning in Large Language Models (Wei et al., 2022)
Least-to-Most Prompting Enables Complex Reasoning (Zhou et al., 2022)
Generated Knowledge Prompting for Commonsense Reasoning (Liu et al., 2021)
Contribute to this collection
Know a great resource? Submit a pull request to add it.
Step-Back Prompting(SBP)
Abstracts to higher-level principles before tackling specific problems
🎯 30-Second Overview
Pattern: Reasoning strategy that steps back to broader principles before tackling specific problems
Why: Improves reasoning by accessing relevant foundational knowledge and applying general principles to specific cases
Key Insight: Specific problem → Ask broader questions → Answer general principles → Apply to specific → Enhanced solution
⚡ Quick Implementation
📋 Do's & Don'ts
🚦 When to Use
Use When
- • Complex problems requiring foundational knowledge
- • Domain-specific questions needing context
- • When initial reasoning attempts fail
- • Problems with multiple solution approaches
- • Educational or explanatory contexts
Avoid When
- • Simple factual queries
- • Well-defined procedural tasks
- • When direct knowledge access is sufficient
- • Time-sensitive quick decisions
- • Problems with obvious solution paths
📊 Key Metrics
💡 Top Use Cases
References & Further Reading
Deepen your understanding with these curated resources
Academic Papers
Take a Step Back: Evoking Reasoning via Abstraction in Large Language Models (Zheng et al., 2023)
Chain-of-Thought Prompting Elicits Reasoning in Large Language Models (Wei et al., 2022)
Least-to-Most Prompting Enables Complex Reasoning (Zhou et al., 2022)
Generated Knowledge Prompting for Commonsense Reasoning (Liu et al., 2021)
Contribute to this collection
Know a great resource? Submit a pull request to add it.