跳到主要内容Skip to main content

← 返回AI 论文← Back to AI Papers

推理Reasoning

ReAct: Synergizing Reasoning and Acting in Language ModelsReAct: Synergizing Reasoning and Acting in Language Models

问题:推理(思维链)与行动(动作规划)此前被分开研究,能不能让模型交替「想一步、做一步、看结果」再继续。
证据位置:摘要给出 ALFWorld 与 WebShop 分别高出模仿学习与强化学习 34% 与 10% 的绝对成功率;HotpotQA 27.4、FEVER 60.9 在表 1。
作者自陈局限:没有独立的 Limitations 章节,边界散在 §3.3 与 §6——交替结构提高了可追溯性,但也限制了推理自由度,推理错误率比纯思维链更高;检索到的信息是否有用是成败关键,无效检索占错误案例的 23%;大动作空间任务需要更多示例,很容易超出上下文长度上限;作者还指出所有提示方法都仍显著落后于领域专用的 SOTA,且离人类专家水平很远。
它没主张的:论文没有说 ReAct 在所有任务上都优于思维链——在 HotpotQA 上它比 CoT 略低(27.4 对 29.4),只在 FEVER 上更高;也没有说它能替代微调或达到人类水平。
Problem: reasoning (chain of thought) and acting (action planning) had been studied separately; can a model interleave thinking, acting and observing.
Where the evidence sits: the abstract gives absolute success-rate gains of 34% and 10% over imitation and reinforcement learning on ALFWorld and WebShop; HotpotQA 27.4 and FEVER 60.9 are in Table 1.
What the authors say about limits: no separate Limitations section; the boundaries are in 3.3 and 6 — the interleaved structure improves groundedness and trustworthiness but constrains reasoning flexibility, raising the reasoning error rate above plain chain of thought; whether retrieval returns useful knowledge is critical, and non-informative search accounts for 23% of error cases; tasks with large action spaces need more demonstrations, which easily exceed the in-context length limit; and all prompting methods remain well behind domain-specific state of the art and far from expert humans.
What it does not claim: the paper does not say ReAct beats chain of thought on every task — on HotpotQA it is slightly lower (27.4 versus 29.4) and only higher on FEVER — nor that it replaces fine-tuning or reaches human level.

关键参数Key specs

证据位置 / evidence
摘要;表 1(HotpotQA / FEVER);§3.3 与 §6
ALFWorld
最佳 ReAct 平均成功率 71%(Act 45%、BUTLER 37%)
WebShop
成功率绝对提升 10%
HotpotQA / FEVER
27.4 EM / 60.9 Acc
无效检索占比 / failed retrieval
错误案例的 23%
复现状态 / reproduction
未复现

官方原文怎么说What the source says

While large language models (LLMs) have demonstrated impressive performance 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.
1 https://ar5iv.labs.arxiv.org/html/2210.03629
While interleaving reasoning, action and observation steps improves ReAct’s groundedness and trustworthiness, such a structural constraint also reduces its flexibility in formulating reasoning steps, leading to more reasoning error rate than CoT.
2 https://ar5iv.labs.arxiv.org/html/2210.03629
For ReAct, successfully retrieving informative knowledge via search is critical. Non-informative search, which counts for 23% of the error cases, derails the model reasoning and gives it a hard time to recover and reformulate thoughts.
3 https://ar5iv.labs.arxiv.org/html/2210.03629
Despite the simplicity of our method, complex tasks with large action spaces require more demonstrations to learn well, which unfortunately can easily go beyond the input length limit of in-context learning.
4 https://ar5iv.labs.arxiv.org/html/2210.03629

资料与许可Source and licence

资料时间Source checked
2026-11-07
许可 / 类型Licence / kind
arxiv-preprint
许可说明Licence evidence
https://arxiv.org/abs/2210.03629
类目Category
推理Reasoning

论文里的数字怎么读How to read the numbers

先分清测什么、怎么算、和谁比、有没有数据污染。拿不准就回原文看实验设置。Before citing a score, separate what it measures, how it is scored, what it compares against and whether contamination is controlled. If unsure, go back to the experiment setup in the paper. 看评测基准See benchmark cards · 看数据集卡See dataset cards

去官网看原页Open the official page →

← 返回AI 论文← Back to AI Papers