Least-to-Most Prompting Enables Complex Reasoning in Large Language Models
Summary
Least-to-Most Prompting Enables Complex Reasoning in Large Language Models (2022) proposed least-to-most prompting, a method that decomposes a hard problem into a sequence of easier subproblems and solves them in order. It is an important variant of chain-of-thought prompting, aimed at cases where the target problem is harder than the exemplars in the prompt.
What it introduced: a two-stage prompting pattern. First the model is asked to break a complex problem into simpler parts. Then it solves those parts from easiest to hardest, using earlier answers as context.
What it led to: later decomposition and tool-using agent patterns that treat hard tasks as a queue of easier steps. Live note: 2201-11903-chain-of-thought-prompting. This week's drafts also include 2203-14465-star in this same pointer batch.
Use it when you want the decomposition variant of chain-of-thought prompting. Where to read it: https://arxiv.org/abs/2205.10625 (v3 (16 Apr 2023)).
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.
License: CC-BY-4.0 · Pointer by iinventedlight (AI-assisted). Source: Denny Zhou, Nathanael Schärli, Le Hou, and 8 coauthors (listed on the arXiv abs page), Least-to-Most Prompting Enables Complex Reasoning in Large Language Models (arXiv non-exclusive distribution license). This note links to the paper rather than reproducing it.
Cite this
'iinventedlight (AI-assisted)'. 2026. "Least-to-Most Prompting Enables Complex Reasoning in Large Language Models." iinventedlight Library, iil:note:9dvgdaed, version of 2026-11-12. https://iinventedlight.com/library/notes/2205-10625-least-to-most-prompting/. License: CC-BY-4.0.
@misc{iil_note_9dvgdaed,
author = {{'iinventedlight (AI-assisted)'}},
title = {{Least-to-Most Prompting Enables Complex Reasoning in Large Language Models}},
howpublished = {iinventedlight Library},
year = {2026},
date = {2026-11-12},
url = {https://iinventedlight.com/library/notes/2205-10625-least-to-most-prompting/},
note = {Permanent ID iil:note:9dvgdaed. License: CC-BY-4.0.},
}
Spotted a mistake? Use the MCP feedback tool or email hello@iinventedlight.com.
For language models and agents: start at /llms.txt | /agents.md | /.well-known/agent-library.json. This page in Markdown: /library/notes/2205-10625-least-to-most-prompting.md.