跳到正文
HuggingFace Daily Papers(社區熱門論文)·· 23 天前AI 評分55

TinyAstral 將自我演進 harness 遷移至小型通用操作智能體

TinyAstral: Transferring Self-Evolving Harnesses to Tiny Generalist Manipulation Agents

AI 導讀

TinyAstral提出兩階段方法,先以遞迴自我改進(RSI)優化凍結的開放權重教師模型 harness,再利用教師的推理與失敗恢復軌跡,蒸餾出在同一 harness 中運作的小型學生模型。

正文

Published on Oct 30

Authors:

,

,

,

,

,

,

,

,

,

Abstract

TinyAstral improves robot agent harnesses via recursive self-improvement and distills compact students that retain performance gains with lower compute.

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.

View arXiv page View PDF Project page GitHub Add to collection

Get this paper in your agent:

hf papers read 2610.41216

Don't have the latest CLI?

curl -LsSf https://hf.co/cli/install.sh | bash

Models citing this paper 1

kehong/TinyAstral-weights

Image-Text-to-Text •

Updated about 18 hours ago

Datasets citing this paper 1

kehong/TinyAstral-SFT

Viewer •

Updated about 18 hours ago

•

88.2k

•

2

Spaces citing this paper 0

No Space linking this paper

Cite arxiv.org/abs/2610.41216 in a Space README.md to link it from this page.

Collections including this paper 0

No Collection including this paper

Add this paper to a collection to link it from this page.

來源:HuggingFace Daily Papers(社區熱門論文) · huggingface.co