---
id: 20261016-T0-11
title: "ClawLoop：让AI智能体在训练循环中自我改进"
title_en: "ClawLoop Enables Self-Improving LLM Agents via Native Training Loop"
url: https://ai.daily.yangsir.net/daily/20261016-T0-11
issue_date: 2026-10-16
publish_date: 2026-10-15T04:00:00.000Z
category: research
source_name: "arXiv cs.AI"
source_url: https://arxiv.org/abs/2610.14118
---

# ClawLoop：让AI智能体在训练循环中自我改进

arXiv发布论文ClawLoop，提出一种让LLM智能体在沙箱环境中通过原生训练循环实现自我改进的方法。当前LLM智能体已能直接操作文件系统和Shell，但如何让其在执行中持续优化自身能力仍是难题。ClawLoop通过在harness层面嵌入训练循环，使智能体能在任务执行过程中不断调整策略。该方法面向通用型智能体场景，有望减少人工调参需求，让开发者在构建自动化工作流时获得更少人工干预的持续优化能力。

## English Version

**ClawLoop Enables Self-Improving LLM Agents via Native Training Loop**

arXiv paper ClawLoop proposes a harness-native training loop that enables LLM agents to self-improve while operating in sandbox workspaces with direct file system and shell access. As agents increasingly run in general-purpose harness environments, the challenge of continuous capability improvement without manual intervention remains. ClawLoop embeds the training loop directly into the agent harness, allowing iterative refinement during task execution. This approach could reduce manual tuning for developers building autonomous workflows, enabling agents that get better at their jobs over time.

---

**来源**：[arXiv cs.AI](https://arxiv.org/abs/2610.14118)

**详情页**：https://ai.daily.yangsir.net/daily/20261016-T0-11

---

*智语观潮 · Daily — https://ai.daily.yangsir.net/llms.txt*