Self-Consistency Improves Chain of Thought Reasoning in Language Models
Authors: Xuezhi Wang, Jason Wei, Dale Schuurmans, Quoc Le, Ed Chi, Sharan Narang, Aakanksha Chowdhery, Denny Zhou
Published March 21, 2022
Read the paper: arXiv 2203.11171 · PDF
Why it matters
In 2026 with fast / cheap inference, self-consistency is a free quality lever for any task with a discrete answer. Classification, extraction, yes/no -- sample 5x, vote, done. The only reasons not to use it are latency budgets and hard single-sample determinism requirements.
How to use it
Implement as a wrapper over your normal call path: run the same prompt N times at temperature=0.7, tally answers, return the plurality. For production use, pair with a confidence check -- if the plurality is thin, fall back to a smarter model.
Abstract
We propose a simple decoding strategy -- self-consistency -- which samples diverse reasoning paths and selects the most consistent answer by taking a majority vote.
Articles that use this
- Delete Your Temperature Parameter: Why Self-Consistency Sampling Is Dead on Reasoning Models Like Gemini 3.5 FlashTechniques
- Agent Outcome Rubrics Are Just Prompts: How to Write Grading Criteria That Actually Fail Bad OutputAgentic
- When to Use a Reasoning Model, and When Thinking Just Costs MoreTechniques
- XML Tags Are the New Markdown: Why Closed Delimiters Beat Free-Form PromptsTechniques
- Chain-of-Thought Is Sabotaging Your Model's Sense of Direction: Prompt Grids as Symbols, Not SentencesTechniques

