Self-correction (LLM)
Self-correction is a prompting pattern where the model reviews its own initial answer, identifies errors, and produces a revised answer — a cheap reliability boost for many tasks.
Self-correction (sometimes called "critique-then-revise") gives the model a second pass at its own work. Variants: pure self-critique (same model reviews itself), critic-then-actor (a separate model critiques, then the original revises), constitutional self-correction (rules-based critique). The technique works on factual accuracy, code correctness, and tone alignment with mixed empirical results — strong models gain less than weaker models. In 2026 reasoning models (o-series, Claude with extended thinking) effectively self-correct internally during thinking, so explicit self-correction is most valuable on smaller non-reasoning models.
When to use self-correction (llm)
- Non-reasoning models on hard tasks.
- Code generation where compile-then-fix loops are feasible.
- Factual generation with verification budget.
Common mistakes
- Self-correction with the same prompt — model rationalises its first answer.
- Over-applying on strong models — adds cost with no quality gain.
FAQ
What is self-correction (llm)?
Self-correction is a prompting pattern where the model reviews its own initial answer, identifies errors, and produces a revised answer — a cheap reliability boost for many tasks.
When should I use self-correction (llm)?
Non-reasoning models on hard tasks. Code generation where compile-then-fix loops are feasible. Factual generation with verification budget.
What are the most common mistakes with self-correction (llm)?
Self-correction with the same prompt — model rationalises its first answer. Over-applying on strong models — adds cost with no quality gain.
Related terms
- Chain-of-verification — Chain-of-verification (CoVe) is a prompting technique where the model first drafts an answer, then generates verification questions for each claim, answers them independently, and revises the draft accordingly.
- Chain-of-thought prompting — Chain-of-thought (CoT) prompting tells a language model to write its reasoning steps before its final answer, increasing accuracy on multi-step problems.
- Self-consistency — Self-consistency runs the same prompt multiple times at non-zero temperature and picks the most common final answer, raising accuracy on reasoning tasks.
- Reasoning model — A reasoning model is an LLM trained to produce extensive internal chain-of-thought before its final answer, trading latency for higher accuracy on hard problems.
Last updated: 2026-06-01. Raw markdown: https://promtable.com/glossary/self-correction.md.