Advanced Reasoning Strategies / Reasoning Decomposition

Least-to-Most Prompting

Advanced [4/5]
Progressive problem solving Scaffolded reasoning

Definition

Least-to-Most prompting breaks complex problems into simpler subproblems, solving each sequentially from easiest to hardest. Each solution builds on previous ones, creating a scaffold that makes the final problem tractable.

This technique is particularly effective for problems that have natural hierarchical decomposition.

Key Concepts

  • Problem decomposition: Breaking down into ordered subproblems
  • Progressive complexity: Starting simple, building to complex
  • Solution chaining: Each answer feeds into the next
  • Scaffolded learning: Building understanding step by step

Examples

Math Problem
Building Block Approach
Original: "How many ways can 5 people sit in a row if Alice and Bob must sit together?" Decomposition: Q1 (simplest): "How many ways can 2 people sit together?" A1: 2 ways (AB or BA) Q2: "If we treat Alice-Bob as one unit, how many units total?" A2: 4 units (Alice-Bob + 3 others) Q3: "How many ways can 4 units be arranged?" A3: 4! = 24 ways Q4 (original): "Combine previous answers" A4: 2 × 24 = 48 ways
Two-Phase Process
Decompose Then Solve
Phase 1 - Decomposition Prompt: "Break this problem into simpler subproblems that can be solved sequentially: [complex problem]" Phase 2 - Sequential Solving: "Solve subproblem 1: [simplest problem]" "Using that answer, solve subproblem 2: [next problem]" ... "Using all previous answers, solve the final problem"

Interactive Exercise

Decompose a Problem

Break this into least-to-most subproblems:

Problem: "Calculate the total cost of a road trip: 500 miles, car gets 25 mpg, gas costs $4/gallon, plus $50/night hotel for 2 nights."

Pro Tips
  • Start with the absolute simplest subproblem
  • Each step should clearly build on previous answers
  • Verify intermediate answers before proceeding
  • Works especially well for math and logic problems

Related Terms