# ysymyth/ReAct

[ICLR 2023] ReAct: Synergizing Reasoning and Acting in Language Models

Repository: https://github.com/ysymyth/ReAct
Canonical: https://ross.abutalabs.com/products/ysymyth-react
Language: Jupyter Notebook
License: MIT
License Family: permissive
Topics: decision-making, large-language-models, llm, prompting, reasoning
Last push: 2024-02-06T02:34:32+00:00

## Health v2 (maintenance only)
Score: 32/100 (v2, computed 2026-09-03T02:39:23.370411+00:00)
- activity 0, release rhythm 35, longevity 99
- inputs: {"age_days": 1389, "days_push": 940, "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 4122, forks 398 (observed 2026-08-28T04:08:35.984093+00:00)

## What it is
Reference prompting code for the ICLR 2023 ReAct paper, which synergizes reasoning and acting in language models. It provides Jupyter notebooks reproducing experiments on HotpotQA, FEVER, AlfWorld, and WebShop using GPT-3 and PaLM.

## Use cases
- reproduce ReAct paper experiments on question answering and fact verification
- learn how to prompt LLMs to interleave reasoning traces with actions
- evaluate GPT-3 on HotpotQA, FEVER, AlfWorld, and WebShop benchmarks
- understand the ReAct prompting pattern before building agents with LangChain
- study prompting strategies for LLM decision-making

## When to choose
- you want the original reference implementation of ReAct prompting
- you are reproducing the paper's benchmark results
- you are learning how reasoning-plus-acting prompts work

## When to avoid
- you need a production-ready agent framework - use LangChain or similar instead
- you want to run ReAct on tasks beyond the four benchmarked notebooks without writing your own code
- you need support for models other than OpenAI GPT-3 or PaLM

## Facets
- artifact type: learning-resource
- maturity: stable
- function: prompt-engineering, agent-framework, llm-inference
- domain: large-language-models, artificial-intelligence
- platform: python
- tags: react-prompting, research-paper, jupyter-notebooks, reasoning-and-acting, gpt-3, ai-agents

## Member repositories
- ysymyth/ReAct (main) score 32

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:08:35.984093+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 2026-08-29T18:23:06.476084+00:00, confidence not recorded.
  - readme: https://github.com/ysymyth/ReAct (fetched 2026-08-28T04:08:35.984093+00:00, sha bdc517929b01)
- Data as of 2026-08-30T08:39:29.467469+00:00.