Comparison
Chain-of-Thought vs Self-Consistency
Chain-of-Thought and Self-Consistency are both common AI/LLM terms but cover different ideas. Here is a quick side-by-side.
When you would reach for Chain-of-Thought
Chain-of-Thought comes up when the question is fundamentally about prompting.
"Solve this word problem step by step." — model shows working.
When you would reach for Self-Consistency
When the task has a verifiable answer (math, logic, code that compiles) and N× compute is acceptable.
A GSM8K eval: sample 32 CoT completions per problem, take the majority numeric answer.
Frequently asked
What is the difference between Chain-of-Thought and Self-Consistency?
Chain-of-Thought: Chain-of-thought prompting asks the model to show its reasoning step by step before giving a final answer. It dramatically improves performance on multi-step problems. Self-Consistency: Self-consistency samples N chain-of-thought completions for the same problem and takes the majority answer. Improves accuracy on math and reasoning tasks at N× the cost.
When should I use Chain-of-Thought vs Self-Consistency?
Chain-of-Thought is the right concept when you are focused on prompting. When the task has a verifiable answer (math, logic, code that compiles) and N× compute is acceptable.
Are Chain-of-Thought and Self-Consistency the same thing?
No. Chain-of-Thought is prompting; Self-Consistency is prompting. They are related but address different parts of the AI stack.