“For example, take a look at the ReAct paper: ReAct”
ReAct: Synergizing Reasoning and Acting in Language Models
让 LLM 交替生成推理轨迹与工具调用
提出 ReAct:让 LLM 交替生成推理轨迹和任务相关的行动,推理帮助模型制定、跟踪和更新行动计划并处理异常,行动则让它借助知识库或环境获取额外信息。在 HotpotQA 和 Fever 上,ReAct 通过调用简单的 Wikipedia API,缓解了思维链中常见的幻觉和误差传播;在 ALFWorld 和 WebShop 上,仅用一两个上下文示例,成功率分别比模仿学习和强化学习方法高 34% 和 10%(绝对值)。
评论者把它当作 LLM agent 的基础范式:模型先说明意图再调用工具,被 LangChain 等框架采用,并作为构建自主 agent 和调试 agent 行为时的原始出处。
HN 书评
推荐在前,同类按时间 · 每篇最多保留 20 条,统计用全部提及“To get a sense of what is and will be possible, take a look at the ReAct paper: ReAct”
“For example, take a look at the ReAct paper: ReAct”
“Implementing the ideas in the ReAct paper, which I have done, is enough to give an AI volition and agency, for any useful definition of the terms.”
“The original 2022 ReACT paper is still best explainer: ReAct”
“Also it seems appropriate to cite the original ReAct paper (from Google mainly)”
看全部 19 条 →
“We use the ReAct paradigm so it explains what it's trying to do before taking an action, which both makes it more accurate and helps with debugging.”
“it's all based on the ReAct paper[0]”
“A lot of this can be solved by using clever prompting and allowing the LLM to use external tools - see the ReAct paper ReAct”
“You could argue being able to observe and think about what was just said aligns with ReAct”
“which is using the REACT model [0] so it can run code, communicate with other apps, etc.”
“The paper author also wrote a paper called ReAct [2], which is one of the core components of the langchain framework.”
“Sure it can. It's not default behavior sure but it's fairly trivial to set up, just expensive.”
“ReAct: ReAct”
“White paper for reference:”
“it appears to be about a Reasoning and Acting methodology[0] with LLMs.”
“Prompts are handled using a ReAct[1] agent which has access to the schema and a tool for executing queries.”
“Is this not a variation of ReAct + Chain-of-Thought + Structured Planning?”
“he's just talking about going from ReAct [1] to Ratchet [2].”
- ReAct: Synergizing Reasoning and Acting in Language Models ↗ScottStevenson · 2023-01-01 · 1 分