# amazon-science/auto-cot

Official implementation for "Automatic Chain of Thought Prompting in Large Language Models" (stay tuned & more will be updated)

Repository: https://github.com/amazon-science/auto-cot
Canonical: https://ross.abutalabs.com/products/auto-cot
Homepage: https://arxiv.org/abs/2210.03493
Language: Jupyter Notebook
License: Apache-2.0
License Family: permissive
Topics: large-language-models, prompt-engineering, gpt-3, gpt3-prompts, gpt3-resources, reasoning, chain-of-thought
Last push: 2024-03-13T12:25:02+00:00

## Health v2 (maintenance only)
Score: 32/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 0, release rhythm 35, longevity 100
- inputs: {"age_days": 1428, "days_push": 903, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 2046, forks 192 (observed 2026-08-28T04:06:08.885835+00:00)

## What it is
Official implementation of Auto-CoT (ICLR 2023), a method that automatically constructs chain-of-thought demonstrations for large language models by sampling diverse questions and generating reasoning chains. It eliminates manual prompt demonstration design while matching or exceeding manual CoT performance on GPT-3 across ten benchmark reasoning tasks.

## Use cases
- automatically build chain-of-thought prompt demonstrations for GPT-3
- avoid hand-crafting few-shot reasoning examples for LLM prompting
- reproduce the Auto-CoT ICLR 2023 paper results
- improve LLM accuracy on multi-step reasoning benchmarks like MultiArith
- generate diverse question clusters with reasoning chains for prompting
- compare zero-shot CoT vs automatic few-shot CoT prompting

## When to choose
- you need chain-of-thought demonstrations without manual prompt engineering
- you are doing research on LLM prompting and reasoning chains
- you want to replicate or extend the Auto-CoT paper on GPT-3

## When to avoid
- you need a production LLM application framework with APIs for modern models
- you want maintained tooling - the repo is a research artifact with limited updates
- you work with models other than GPT-3-era APIs without adaptation

## Facets
- artifact type: library
- maturity: maintenance
- function: prompt-engineering, machine-learning, nlp
- domain: large-language-models, artificial-intelligence, machine-learning
- platform: python
- tags: chain-of-thought, gpt-3, llm-prompting, reasoning, research-code, iclr-2023, natural-language-processing

## Member repositories
- amazon-science/auto-cot (main) score 32

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:06:08.885835+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T02:57:59.421273+00:00, confidence not recorded.
  - readme: https://github.com/amazon-science/auto-cot (fetched 2026-08-28T04:06:08.885835+00:00, sha 72e2a633d5fc)
  - homepage: https://arxiv.org/abs/2210.03493 (fetched 2026-08-29T10:38:16.328615+00:00, sha 701c5492d1e8)
  - site_page: https://info.arxiv.org/about/donate.html (fetched 2026-08-29T10:38:16.338366+00:00, sha cca9c3a11c56)
  - site_page: https://info.arxiv.org/about/ourmembers.html (fetched 2026-08-29T10:38:16.341864+00:00, sha 47cbc55ff1de)
  - site_page: https://info.arxiv.org/about (fetched 2026-08-29T10:38:16.343568+00:00, sha a1f16f915a9a)
  - site_page: https://info.arxiv.org/labs/index.html (fetched 2026-08-29T10:38:16.340177+00:00, sha b14a8d05a0ec)
- Data as of 2026-08-30T08:39:29.467469+00:00.