论文HF Daily Papers · arXiv 2610.412164 天前

TinyAstral：4B模型蒸馏实现62.8%操作成功率与5倍推理加速

提出TinyAstral两阶段框架，通过递归自改进优化教师Harness使成功率从33.4%升至63.6%，蒸馏出的4B学生模型达62.8%成功率，加速器耗时仅为教师五分之一，验证了系统级优化向小模型迁移的有效性。

AI 整理

具身模型
端侧部署

本文由 AI 根据公开内容整理，可能有误，请以原文为准：TinyAstral: Transferring Self-Evolving Harnesses to Tiny Generalist Manipulation Agents（HF Daily Papers · arXiv 2610.41216）

核心内容

针对具身智能体依赖大模型导致推理成本高的问题，研究提出TinyAstral框架。该方法首先在RoboTwin基准上通过递归自改进（RSI）机制优化冻结教师的工具与反馈逻辑，使其任务成功率由33.4%提升至63.6%。随后固定演进后的Harness，利用包含推理与故障恢复的教师轨迹蒸馏出4B参数学生模型。实验显示该学生模型成功率达62.8%，在相同部署场景下加速器占用时间仅为原教师模型的20%，且性能优于SenseNova-RoboRSI等开源竞品。

技术核心在于将系统级能力解耦为可演进的Harness组件。RSI阶段由前沿LLM代理读取教师轨迹并提出编辑建议，仅当通过配对统计检验时才采纳修改，确保优化稳健性。蒸馏阶段保留最终Harness不变，使学生模型继承教师的工具调用、记忆管理及错误恢复逻辑。结果表明，Harness演进收益可无损传递至紧凑模型，且在未参与训练的基准上，4B学生模型表现接近运行相同Harness的552B规划器，证实了“强系统+小模型”路线在通用操作任务中的工程可行性。

关键看点

递归自改进：利用LLM代理自动提议并统计验证Harness编辑，使教师模型在RoboTwin成功率从33.4%跃升至63.6%
高效蒸馏：4B学生模型继承优化后Harness，成功率达62.8%，逼近教师水平且显著优于同级开源Agent
推理加速：在实测部署中，4B学生模型加速器耗时仅为教师模型的20%，大幅降低端侧算力门槛
泛化验证：在未训练基准上，4B模型配合演进Harness性能接近552B规划器，证明系统优化可弥补参数量差距

工程与产业观察

该工作证明了具身智能性能提升不必单纯依赖模型 scaling，通过结构化优化外部Harness并蒸馏至小模型，可在保持高性能的同时实现5倍推理加速。这为端侧部署提供了新范式：将复杂逻辑沉淀于可复用的系统组件，而非全部压缩进权重，对低成本机器人量产具有直接参考价值。

英文摘要 (Abstract)

Robot agents built on vision-language models depend on both the planner and its harness: the tools, feedback, memory, and recovery logic around it. Can improvements to a large teacher's harness be carried into a compact agent with lower inference compute cost? We present TinyAstral, a two-stage approach. We first freeze an open-weight teacher and improve its harness through recursive self-improvement (RSI): frontier LLM agents read the teacher's trajectories and propose harness edits, and each edit is accepted only if it passes a paired statistical test. We then freeze the final harness, collect teacher trajectories with reasoning and failure recovery, and distill a small student that replaces the teacher in the same harness. On RoboTwin, harness evolution raises the teacher's success from 33.4% to 63.6%; the final 4B student reaches 62.8%. Across four harness versions, distilled 4B students score close to their corresponding teachers, and later versions yield stronger students, showing that harness evolution gains carry into compact models. The harness gains also transfer to other frozen planners without retraining. On identical episodes, the final 4B student outperforms open 4B agents, the open 27B harness-RSI model SenseNova-RoboRSI, and our own teacher model placed in three competitors' harnesses. It stays close to its teacher on all three training benchmarks, and on three benchmarks never used for RSI or training it scores close to a 552B planner running the same harness. On the measured deployment episodes, it uses about one fifth of its teacher's accelerator time.

Hugging Face Daily Papers 热度：3 upvotes 作者：Kehong Gong, Xiaoxu Zheng, Guanli Hou, Seunghun Lee, Mengqi Guo, Runjia Tan, Zhengyu Wen, Qi Wang, Xiaoyu He, Mingyuan Zhang, Michael Bi Mi