# RoboArchitect：自动化Agent框架设计使冻结策略成功率提升4.5倍

> 原文链接：https://robotics.hellobike.com/papers/2026-10-30-2610-42141-c17a67a2
> 频道：papers | 更新日期：2026-10-30 | 标签：auto, arxiv, hf-daily-papers
> 核心摘要：提出RoboArchitect系统，通过代码模型自动迭代优化机器人Agent框架而非微调权重。在未见场景与任务中，冻结策略平均成功率分别从9.7%升至43.3%、12.7%升至40.2%，验证了架构自动化设计的巨大潜力。

# RoboArchitect：自动化Agent框架设计使冻结策略成功率提升4.5倍

::: v-pre
> 本文由 AI 根据公开内容整理，可能有误，请以原文为准：[RoboArchitect: Automating the Design of Robotic Agents with Persistent Memory](https://arxiv.org/abs/2610.42141)（HF Daily Papers · arXiv 2610.42141）

## 核心内容

针对机器人策略迁移需手工重构Agent框架的痛点，研究提出RoboArchitect自动化设计方法。该系统利用编码模型观察Rollout数据、诊断故障并迭代修订围绕冻结策略的控制框架。实验显示，在不更新任何模型权重的前提下，该方法在多个基准测试中将未见场景的平均成功率从9.7%提升至43.3%（4.5倍），未见任务成功率从12.7%提升至40.2%（3.2倍），且当基准与训练数据分布差异较大时，演化出的专用框架性能常优于微调版本。

技术层面，RoboArchitect生成的框架具备自适应控制策略与持久化记忆机制。系统能自动习得如“抬起卡死机械臂”或“水平绕障抓取”等直观操作逻辑，并将有效抓取深度等经验存入持久记忆供后续决策调用。这表明机器人操作能力的显著提升空间存在于Agent框架设计中，通过自动化代码生成与优化，可在零权重更新下实现比传统微调更优的泛化效果，为具身智能部署提供了新范式。

## 关键看点

- **零权重更新**：仅通过自动化迭代Agent框架代码，使冻结策略在未见场景成功率提升4.5倍至43.3%
- **持久化记忆**：框架自动演化出记忆模块，存储有效抓取深度等经验以优化后续长程决策
- **自适应控制**：自动生成抬臂解卡、水平绕障等直观策略，解决通用策略在特定场景的失效问题
- **超越微调**：在数据分布偏移显著的场景下，自动化框架封装的通用策略性能优于全量微调模型

## 工程与产业观察

该研究揭示了具身智能落地中被低估的“系统工程红利”。在基础模型权重冻结前提下，通过LLM驱动的Agent框架自动化演进，可低成本实现特定场景适配。这提示产业界在追求模型Scaling Law之外，应重视推理侧代码架构与记忆机制的工程化挖掘，将其作为解决Sim-to-Real鸿沟与长尾问题的关键杠杆。

## 英文摘要 (Abstract)

Recent work shows that a frozen robot policy performs far better inside an agent harness, where control code decides when to query the policy, consult vision tools, or command the arm directly. Yet today these harnesses are built by hand and redesigned for each new robot or benchmark. We show that harness design can be automated, offering a strong alternative to finetuning on a novel setup. In RoboArchitect, a coding model observes a robot's rollouts, diagnoses failures, and iteratively revises the harness around the frozen policy. The harness evolves on a set of episodes for its optimization and is evaluated on held-out scenes and tasks. Without updating any model weights, RoboArchitect raises average success from 9.7% to 43.3% on unseen scenes (4.5x) and from 12.7% to 40.2% on unseen tasks (3.2x), across multiple benchmarks and policies. The gains are largest where the benchmark is far from the policy's training data: wrapping a general-purpose policy, the evolved harness yields a specialist that often outperforms even the finetuned version. Intuitive control strategies are automatically designed, such as lifting a jammed arm clear or reaching in horizontally when an obstacle blocks the way from above. The harness also evolves its own persistent memory, carrying learned experience, such as effective grasp depths, into later decisions. These results suggest that substantial headroom in robotic manipulation lies in the agent harness, and its design can be effectively automated and optimized without weight updates.

Hugging Face Daily Papers 热度：5 upvotes
作者：Shengguang Wu, Suning Huang, Cherie Ho, Ria Doshi, Hao Zhu, Serena Yeung-Levy
:::

## 原始参考来源 (Sources)

- [HF Daily Papers · arXiv 2610.42141：RoboArchitect: Automating the Design of Robotic Agents with Persistent Memory](https://arxiv.org/abs/2610.42141)