---
title: 'STaR: Bootstrapping Reasoning With Reasoning'
license: CC-BY-4.0
license_url: https://creativecommons.org/licenses/by/4.0/
license_note: 'paper is arXiv non-exclusive distribution license (verified on arXiv abs page 2026-11-12); this pointer note is our own words and is CC BY 4.0; we link to the paper rather than reproduce it'
source: https://iinventedlight.com/library/notes/2203-14465-star.md
source_license: 'arXiv non-exclusive distribution license'
attribution: 'Pointer by iinventedlight (AI-assisted). Source: Eric Zelikman, Yuhuai Wu, Jesse Mu, Noah D. Goodman, STaR: Bootstrapping Reasoning With Reasoning (arXiv non-exclusive distribution license). This note links to the paper rather than reproducing it.'
author: 'iinventedlight (AI-assisted)'
date: 2026-11-12
status: published
approved_by: "Cole Withington, 2026-11-12 14:44 PT, Librarian chat widget t51s316: \"Yes, publish all 100 second-list arXiv notes\""
checked_on: 2026-11-12
as_of: 2026-11-12
confidence: medium
untested: 'Pointer note only. Paper not reproduced. Abs page and license checked 2026-11-12.'
category: arxiv
topic_area: 'ai-research: reasoning'
tags: arxiv, reasoning, self-training, chain-of-thought
arxiv_id: 2203.14465
alias: 2203.14465-star
type: pointer
format: pointer
sources:
  - id: S1
    title: 'STaR: Bootstrapping Reasoning With Reasoning'
    publisher: arXiv
    authors: 'Eric Zelikman, Yuhuai Wu, Jesse Mu, Noah D. Goodman'
    url: https://arxiv.org/abs/2203.14465
    version_read: 'v2 (20 May 2022)'
    accessed: 2026-11-12
    license: 'arXiv non-exclusive distribution license'
---
# STaR: Bootstrapping Reasoning With Reasoning

## Summary
[STaR: Bootstrapping Reasoning With Reasoning](https://arxiv.org/abs/2203.14465) (2022) proposed STaR, a method that bootstraps reasoning by iteratively keeping self-generated rationales that lead to correct answers and then finetuning on those traces. It is an important precursor of later self-taught reasoners.

What it introduced: a loop in which a model generates rationales, those that reach a correct answer are kept, and the model is then trained on that growing set, stretching a small amount of rationale supervision further than a fixed few-shot prompt.

What it led to: later self-improvement and rationale-training recipes. Live notes: [2201-11903-chain-of-thought-prompting](https://iinventedlight.com/library/notes/2201-11903-chain-of-thought-prompting/) and [2408-03314-test-time-compute](https://iinventedlight.com/library/notes/2408-03314-test-time-compute/). This week's drafts also include [2412-16720-o1-system-card](https://iinventedlight.com/library/notes/2412-16720-o1-system-card/) and [2501-12948-deepseek-r1](https://iinventedlight.com/library/notes/2501-12948-deepseek-r1/).

Use it when you want the original rationale-bootstrapping method, before later process-reward or test-time reasoners. Where to read it: https://arxiv.org/abs/2203.14465 (v2 (20 May 2022)).

The paper is posted on arXiv under the arXiv non-exclusive distribution license. This note links to it rather than reproducing it. The note itself is CC BY 4.0.

License: this note is CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). Pointer by iinventedlight (AI-assisted). Paper license: arXiv non-exclusive distribution; this note links to it and does not reproduce it.

---

Agents: [/agents.md](/agents.md) | [/.well-known/agent-library.json](/.well-known/agent-library.json).
