菜单

关于 🐙 GitHub
arXiv 提交日期: 2026-01-22
📄 Abstract - LLM-in-Sandbox Elicits General Agentic Intelligence

We introduce LLM-in-Sandbox, enabling LLMs to explore within a code sandbox (i.e., a virtual computer), to elicit general intelligence in non-code domains. We first demonstrate that strong LLMs, without additional training, exhibit generalization capabilities to leverage the code sandbox for non-code tasks. For example, LLMs spontaneously access external resources to acquire new knowledge, leverage the file system to handle long contexts, and execute scripts to satisfy formatting requirements. We further show that these agentic capabilities can be enhanced through LLM-in-Sandbox Reinforcement Learning (LLM-in-Sandbox-RL), which uses only non-agentic data to train models for sandbox exploration. Experiments demonstrate that LLM-in-Sandbox, in both training-free and post-trained settings, achieves robust generalization spanning mathematics, physics, chemistry, biomedicine, long-context understanding, and instruction following. Finally, we analyze LLM-in-Sandbox's efficiency from computational and system perspectives, and open-source it as a Python package to facilitate real-world deployment.

顶级标签: llm agents systems
详细标签: code sandbox agent framework reinforcement learning tool usage general intelligence 或 搜索:

LLM-in-Sandbox:通过代码沙箱解锁大语言模型的通用智能 / LLM-in-Sandbox Elicits General Agentic Intelligence


1️⃣ 一句话总结

本文提出了LLM-in-Sandbox框架,让大语言模型在代码沙箱(虚拟计算机)中自主探索,无需额外训练即可激发其在数学、物理、化学、生物医学等非代码领域的通用智能,并通过强化学习(LLM-in-Sandbox-RL)进一步提升模型能力。


2️⃣ 论文创新点

1. LLM-in-Sandbox通用框架

2. 共享沙箱环境与最小化工具集设计

3. 基于ReAct的工作流程与基于文件的输入输出处理

4. LLM-in-Sandbox强化学习(LLM-in-Sandbox-RL)


3️⃣ 主要结果与价值

结果亮点

实际价值


4️⃣ 术语表

源自 arXiv: 2601.16206