外观
论文HF Daily Papers · arXiv 2610.295607 天前
SERA框架通过自进化Harness提升VLA长程操作成功率至51.38%
本文由 AI 根据公开内容整理，可能有误，请以原文为准：From VLM Planners to Robotic Agents: Self-Evolving Harnesses for History-Dependent Manipulation（HF Daily Papers · arXiv 2610.29560）
核心内容
针对视觉-语言-动作（VLA）模型在非马尔可夫长程操作任务中因部分可观测性和执行误差累积导致的性能瓶颈，研究提出SERA（Self-Evolving Robotic Agent）框架。该方案将高层VLM组织为具备可编辑Harness的机器人代理，固定底层VLA执行器，通过观测与反馈动态维护任务上下文。在RoboMME基准测试中，SERA经自主配置选择后整体成功率达51.38%，显著优于未启用自进化机制的44.38%基线水平。
SERA架构核心在于分离规划与执行：底层由固定VLA模型负责低层控制，上层Harness作为状态机维护任务进度与交互历史。自进化机制利用环境Rollout数据自动提议并评估Harness修订方案，筛选最优配置用于后续推理，无需微调模型权重即可适配历史依赖型任务。实验表明，这种基于Agent化规划的范式能有效缓解长程操作中的上下文丢失问题，为具身智能体在复杂非结构化场景下的鲁棒决策提供了新的工程路径。
关键看点
- SERA架构：采用固定VLA执行器+可编辑Harness分层设计，通过观测反馈动态更新任务上下文以解决非马尔可夫依赖难题
- 自进化机制：利用环境Rollout自动提议并评估Harness配置修订，实现规划逻辑的自主优化而无需重训底层模型权重
- 性能验证：在RoboMME基准上整体成功率达51.38%，较无自进化基线提升7%，证实Agentic规划对历史依赖操作的有效性
工程与产业观察
SERA验证了在不微调VLA权重前提下，通过外挂可进化记忆模块解决长程操作历史依赖问题的可行性。这为具身智能落地提供了轻量化工程思路：当基础模型能力触顶时，通过Agent化架构设计与运行时自适应优化，可低成本提升复杂任务鲁棒性，降低对海量长程示教数据的依赖。
英文摘要 (Abstract)
Vision-language-action (VLA) models achieve strong performance across embodied manipulation tasks by unifying visual understanding, language conditioning, and action generation. However, long-horizon tasks can involve non-Markovian dependencies: successful decisions may require retaining task progress and past interaction outcomes, while partial observability and execution errors compound over time. Existing work has begun to address these challenges by augmenting foundation-model policies with memory and by introducing hierarchical architectures that separate high-level planning from low-level execution. We organize the high-level VLM as a robotic agent through an editable harness to maintain task context and improve planning through execution experience. We instantiate this approach as SERA (Self-Evolving Robotic Agent), with a fixed VLA executor providing low-level control. The harness maintains and updates task context using observations and execution feedback to guide subsequent planning. Self-evolution uses environment rollouts to propose and evaluate harness revisions, with the selected configuration fixed before held-out testing. On RoboMME, SERA achieves 51.38% overall success with autonomous selection, compared with 44.38% without self-evolution. Our exploration suggests that agentic planning is a promising direction for history-dependent robotic manipulation.
Hugging Face Daily Papers 热度：5 upvotes 作者：Tingwen Du, Haoyu Guo, Hanxue Zhang, Xuanxu Wang, Junting Dong, Chao Gao, Mingfei Han, Tao Lu, Mulin Yu, Chunhua Shen, Xiaojun Chang, Zhihui Li
来源
本页事实对应的公开页面，以原文为准。