Unifying Embodied Foundation Models via a Model-to-Policy Formulation

原文：Unifying Embodied Foundation Models via a Model-to-Policy Formulation

作者：Xinyi Hong, Mingyang Ren

来源：arXiv cs.RO（机器人）

正文

Computer Science > Robotics

arXiv:2610.23053v1 (cs)

[Submitted on 16 Oct 2026]

Title:Unifying Embodied Foundation Models via a Model-to-Policy Formulation

Authors:Xinyi Hong, Mingyang Ren

View a PDF of the paper titled Unifying Embodied Foundation Models via a Model-to-Policy Formulation, by Xinyi Hong and 1 other authors

View PDF

HTML (experimental)

Abstract:Vision-language-action (VLA) models and world-action models (WAMs) have emerged as promising approaches to robot learning and control. These models integrate multimodal perception and action generation, but differ in their architectures, output representations, and learning objectives. In this paper, we present a unified, architecture-agnostic probabilistic formulation for embodied foundation models based on stochastic kernels. Our formulation has two key components: (i) we characterize embodied foundation models through conditional output distributions and action-realization kernels, which together induce task-conditioned robot policies; and (ii) we connect pretraining and post-training to policy deployment, accommodating model-specific pretraining objectives and offline or online post-training strategies. We instantiate the formulation for OpenVLA and DreamZero, covering autoregressive action-token prediction and joint video-action generation.

Subjects:

Robotics (cs.RO); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)

Cite as:

arXiv:2610.23053 [cs.RO]

(or

arXiv:2610.23053v1 [cs.RO] for this version)

https://doi.org/10.48550/arXiv.2610.23053

Focus to learn more

arXiv-issued DOI via DataCite (pending registration)

主题

机器人 · 人工智能 · 机器学习

由「前沿雷达」于 2026-10-16 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。