Learning Robot Behavior Priors from Non-Robotic Synthetic Data

原文：Learning Robot Behavior Priors from Non-Robotic Synthetic Data

作者：Mateusz Odrowaz-Sypniewski, Michael Amir, Jasmine Bayrooti, Ajay Shankar, Amanda Prorok

来源：arXiv cs.RO（机器人）

正文

Computer Science > Robotics

arXiv:2610.31275v1 (cs)

[Submitted on 22 Oct 2026]

Title:Learning Robot Behavior Priors from Non-Robotic Synthetic Data

Authors:Mateusz Odrowaz-Sypniewski, Michael Amir, Jasmine Bayrooti, Ajay Shankar, Amanda Prorok

View a PDF of the paper titled Learning Robot Behavior Priors from Non-Robotic Synthetic Data, by Mateusz Odrowaz-Sypniewski and 4 other authors

View PDF

HTML (experimental)

Abstract:Real robot data, rich across tasks and morphologies, is often treated as the bottleneck for training robot learning models, including embodied foundation models that aim for zero-shot physical transfer. In this paper, we investigate the degree to which such models can acquire necessary structure using data from systems without embodiment. To this end, we test if inexpensive, procedurally generated synthetic data can yield useful priors before a model observes any real-world behavior. We study this through robot action prediction, a natural pretraining objective and a direct test of whether structure learned from arbitrary synthetic tasks transfers to embodied settings. We introduce three data generation pipelines that model mathematical abstractions of robotic control, and use them as pretraining priors for a Transformer. We demonstrate that a model trained purely on this non-robotic data attains non-trivial zero-shot prediction performance on unseen physical robot tasks via an in-context learning mechanism, and that our specialized control priors notably improve upon generic causal priors. We also compare against a compute-matched baseline trained on real robot data, and find that it outperforms our model but only on the datasets it was trained on. These findings suggest that procedurally generated, mathematical abstractions can yield useful robotic priors, offering a potential complement to expensive embodied data collection.

Subjects:

Robotics (cs.RO); Machine Learning (cs.LG)

Cite as:

arXiv:2610.31275 [cs.RO]

(or

arXiv:2610.31275v1 [cs.RO] for this version)

https://doi.org/10.48550/arXiv.2610.31275

Focus to learn more

arXiv-issued DOI via DataCite

主题

机器人 · 机器学习

由「前沿雷达」于 2026-10-22 采集。正文取自原文页面，已保留出处链接。标题与正文版权归原作者所有。