ReAct: Synergizing Reasoning and Acting in Language Models
Shunyu Yao, Jeffrey Zhao, Dian Yu, Nan Du, Izhak Shafran + 2 more
TLDR
ReAct is a method that interleaves reasoning and acting in language models to improve task-solving accuracy and interpretability by enabling interaction with external environments.
Key contributions
- Introduces a unified approach combining reasoning traces and action generation within LLMs for synergistic problem solving.
- Demonstrates improved performance and reduced hallucination on QA and fact verification tasks by interacting with external knowledge sources.
- Achieves significant success rate gains on interactive decision-making benchmarks compared to imitation and reinforcement learning baselines with minimal prompting.
Why it matters
This paper matters because it bridges the gap between reasoning and acting in language models, enabling them to not only think through problems step-by-step but also take meaningful actions to gather information and update plans dynamically. This leads to more reliable, interpretable, and effective AI systems capable of handling complex tasks that require both cognitive reasoning and real-world interaction.
Original Abstract
While large language models (LLMs) have demonstrated impressive capabilities across tasks in language understanding and interactive decision making, their abilities for reasoning (e.g. chain-of-thought prompting) and acting (e.g. action plan generation) have primarily been studied as separate topics. In this paper, we explore the use of LLMs to generate both reasoning traces and task-specific actions in an interleaved manner, allowing for greater synergy between the two: reasoning traces help the model induce, track, and update action plans as well as handle exceptions, while actions allow it to interface with external sources, such as knowledge bases or environments, to gather additional information. We apply our approach, named ReAct, to a diverse set of language and decision making tasks and demonstrate its effectiveness over state-of-the-art baselines, as well as improved human interpretability and trustworthiness over methods without reasoning or acting components. Concretely, on question answering (HotpotQA) and fact verification (Fever), ReAct overcomes issues of hallucination and error propagation prevalent in chain-of-thought reasoning by interacting with a simple Wikipedia API, and generates human-like task-solving trajectories that are more interpretable than baselines without reasoning traces. On two interactive decision making benchmarks (ALFWorld and WebShop), ReAct outperforms imitation and reinforcement learning methods by an absolute success rate of 34% and 10% respectively, while being prompted with only one or two in-context examples. Project site with code: https://react-lm.github.io
📬 Weekly AI Paper Digest
Get the top 10 AI/ML arXiv papers from the week — summarized, scored, and delivered to your inbox every Monday.