Diagram of Thought Prompt Explained: Make AI Think in Connection

Diagram of Thought (DoT) Prompt

Introduction

We’ve seen AI evolve from:
– Thinking step-by-step (Chain of Thought)
– Exploring multiple paths (Tree of Thought)
– Using logic (Logic of Thought)
– Improving itself (Iteration of Thought)

But real human thinking is even more powerful.
– We don’t just think in steps or trees…
– We connect ideas, critique them, refine them, and combine them into a bigger picture. That’s exactly what Diagram of Thought (DoT) prompting brings to AI.

What is Diagram of Thought Prompting?

Think of Diagram of Thought like a mind map
– Ideas are nodes
– Connections are edges
– Some ideas are challenged
– Some are refined
– And finally, everything is combined into a clear answer

In simple terms:
AI doesn’t just think in lines or trees… it thinks in a connected diagram

According to the research paper, DoT:
– Uses a graph structure (DAG) for reasoning
– Includes roles like propose → critique → refine → summarize
– Builds everything within a single model (no external controller needed)

Format of a Diagram of Thought Prompt

A DoT prompt typically follows this flow:

1. Propose Ideas
Generate initial thoughts or solutions

2. Critique
Evaluate each idea:
Is it correct?
Is it weak?

3. Refine
Improve or correct the ideas

4. Validate
Keep only strong, reliable ideas

5. Summarize
Combine all validated ideas into a final answer

In simple terms:
Propose → Critique → Refine → Combine

Let’s See Some Examples!

Writing Improvement Prompt:

Write a blog outline on introducing AI for kids:

Step 1: Generate multiple ideas

Step 2: Critique engagement level

Step 3: Refine structure

Step 4: Combine into final outline

Social Media Marketing Prompt:

You are a strategic thinker using Diagram of Thought reasoning.

Problem: My kids’ newspaper subscription page is getting visitors but no conversions.

Step 1: Propose multiple hypotheses (at least 4–5) for why conversions are low.
Step 2: Critique each hypothesis:

Is it likely or weak?
What evidence would support or reject it?

Step 3: Refine the strongest hypotheses into clearer problem statements.

Step 4: For each refined problem, propose solutions.

Step 5: Critique the solutions:

Impact vs effort
Feasibility for a small team

Step 6: Combine the best ideas into a prioritized action plan.

Step 7: Provide final recommendation with:

Why they will work

Top 3 actions to take immediately

Best Practices

– Encourage interaction between ideas
Don’t treat thoughts independently
– Use critique actively
This is the core strength of DoT
– Focus on refinement
Don’t just generate—improve
– Combine ideas thoughtfully
Final answer should be synthesized

Tips

Diagram of Thought works best for:
– Complex decision making
– Strategy and planning
– Research and analysis
– Multi-step reasoning

It is especially useful when:
You need balanced, well-rounded answers

Think of it like:
“Don’t just think or explore… connect and refine ideas together”

Disclaimer!

LLM like ChatGPT, Gemini can provide incorrect and inaccurate outputs. Always double check the output before you use it.

Have a question?

If you have any other queries, feel free to drop a comment.

Learn More !

Experiment directly with ChatGPT and Google Gemini.
Want to learn more about effective prompts to get the best out of GenAI and LLMs?

Home: Prompt Engineering

Discover more from Debabrata Pruseth

Subscribe to get the latest posts sent to your email.