Prompt EngineeringAdvanced

Branch → Evaluate → Prune

Propose several next steps, judge the partial states, keep the best and expand them. How to build a Tree-of-Thought search, why the evaluator decides everything, and when a reasoning model makes it unnecessary.

By Amit UpadhyayPublished 25 Sep 202610 min

Following one scenario: “Build next week’s on-call rota: nobody twice in a row, Priya off Wednesday, at least one senior every night.”An on-call scheduler for a six-person team with hard constraintsWhy it matters →

Step 1 / 4

Speed

What's happening

Generate multiple thought branches

LLM generates 3 different approaches to the problem

💡 Analogy

Like a chess player mentally exploring three different opening moves before choosing.

Instead of one answer, we ask the model to generate multiple different approaches.

Console Output

> Branch 1: "Use dynamic programming..."
> Branch 2: "Apply greedy algorithm..."
> Branch 3: "Try divide and conquer..."

●tree-of-thought.ts

Generate multiple thought branchesL3–7

1// Tree of Thoughts: explore multiple reasoning paths

2async function treeOfThoughts(problem: string, depth = 3) {

3 // Step 1: Generate candidate thoughts (branches)

4 const thoughts = await llm.complete(`

5 Problem: ${problem}

6 Generate 3 different approaches to solve this:

7 Approach 1:`);

9 // Step 2: Evaluate each thought

10 const scores = await Promise.all(

11 thoughts.map(t => llm.complete(`

12 Rate this approach 1-10 for correctness:

13 Approach: ${t}

14 Score:`))

15 );

17 // Step 3: Keep best branch, expand further

18 const bestThought = thoughts[scores.indexOf(Math.max(...scores))];

20 // Step 4: Recurse on best branch

21 if (depth > 1) return treeOfThoughts(bestThought, depth - 1);

22 return bestThought;

23}

23 lines

Steps · 4

What's happening

Generate multiple thought branches

LLM generates 3 different approaches to the problem

💡 Analogy

Like a chess player mentally exploring three different opening moves before choosing.

Instead of one answer, we ask the model to generate multiple different approaches.

Console Output

> Branch 1: "Use dynamic programming..."
> Branch 2: "Apply greedy algorithm..."
> Branch 3: "Try divide and conquer..."

Next: Test →

tree-of-thought ToT reasoning search beam search prompt-engineering

← PreviousConstrained Generation: Beyond Hope-Based Parsing Next →Red Teaming LLM Applications: Attack It Before Your Users Do

Continue reading

All Prompt Engineering articles →

Prompt Engineering9 min

Meta-Prompting: Letting a Model Tune Your Prompts

From a one-off rewrite to refinement loops on failures, candidate search and compiled DSPy programs. How to optimise prompts against a metric without overfitting, and how to review what the optimiser produces.

Prompt Engineering9 min

Red Teaming LLM Applications: Attack It Before Your Users Do

Threat model first, then manual sessions, then automated scans with garak, promptfoo and PyRIT. How to find jailbreaks, data leaks and unsafe tool calls, fix them in the architecture, and keep them fixed in CI.

Prompt Engineering13 min

Constrained Generation: Beyond Hope-Based Parsing

Enforce output structure at the token level with grammars and token masks — how it differs from native JSON mode, and when the strictness costs you reasoning quality.

Discussion

Prompt EngineeringAdvanced

Branch → Evaluate → Prune

By Amit UpadhyayPublished 25 Sep 202610 min

Step 1 / 4

Speed

What's happening

Generate multiple thought branches

LLM generates 3 different approaches to the problem

💡 Analogy

Like a chess player mentally exploring three different opening moves before choosing.

Instead of one answer, we ask the model to generate multiple different approaches.

Console Output

> Branch 1: "Use dynamic programming..."
> Branch 2: "Apply greedy algorithm..."
> Branch 3: "Try divide and conquer..."

●tree-of-thought.ts

Generate multiple thought branchesL3–7

1// Tree of Thoughts: explore multiple reasoning paths

2async function treeOfThoughts(problem: string, depth = 3) {

3 // Step 1: Generate candidate thoughts (branches)

4 const thoughts = await llm.complete(`

5 Problem: ${problem}

6 Generate 3 different approaches to solve this:

7 Approach 1:`);

9 // Step 2: Evaluate each thought

10 const scores = await Promise.all(

11 thoughts.map(t => llm.complete(`

12 Rate this approach 1-10 for correctness:

13 Approach: ${t}

14 Score:`))

15 );

17 // Step 3: Keep best branch, expand further

18 const bestThought = thoughts[scores.indexOf(Math.max(...scores))];

20 // Step 4: Recurse on best branch

21 if (depth > 1) return treeOfThoughts(bestThought, depth - 1);

22 return bestThought;

23}

23 lines

Steps · 4

What's happening

Generate multiple thought branches

LLM generates 3 different approaches to the problem

💡 Analogy

Like a chess player mentally exploring three different opening moves before choosing.

Instead of one answer, we ask the model to generate multiple different approaches.

Console Output

> Branch 1: "Use dynamic programming..."
> Branch 2: "Apply greedy algorithm..."
> Branch 3: "Try divide and conquer..."

Next: Test →

tree-of-thought ToT reasoning search beam search prompt-engineering

← PreviousConstrained Generation: Beyond Hope-Based Parsing Next →Red Teaming LLM Applications: Attack It Before Your Users Do

Continue reading

All Prompt Engineering articles →

Prompt Engineering9 min