Saved in:
Bibliographic Details
Main Authors: Zhang, Antong, Qi, Han, Yang, Heng
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2605.08571
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866917482972839936
author Zhang, Antong
Qi, Han
Yang, Heng
author_facet Zhang, Antong
Qi, Han
Yang, Heng
contents We introduce BEACON--Best-Effort Adaptation for Cross-Domain Co-Training--a theory-driven framework for training generative robot policies with abundant source demonstrations and limited target demonstrations. BEACON casts cross-domain co-training as a discrepancy-aware importance-reweighting problem, jointly learning a diffusion-based visuomotor policy and per-sample source weights that minimize an objective informed by target-domain generalization guarantees. To make best-effort adaptation practical for high-dimensional sequence policies, we develop scalable instance-level discrepancy estimators, stochastic alternating updates for policy and weights, and a multi-source extension that balances heterogeneous source domains. Across sim-to-sim, sim-to-real, and multi-source manipulation settings, BEACON improves robustness and data efficiency over target-only, fixed-ratio co-training, and feature-alignment baselines. Importantly, even without an explicit alignment objective, BEACON achieves feature alignment as an implicit result of discrepancy-aware cross-domain co-training.
format Preprint
id arxiv_https___arxiv_org_abs_2605_08571
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle BEACON: Cross-Domain Co-Training of Generative Robot Policies via Best-Effort Adaptation
Zhang, Antong
Qi, Han
Yang, Heng
Robotics
We introduce BEACON--Best-Effort Adaptation for Cross-Domain Co-Training--a theory-driven framework for training generative robot policies with abundant source demonstrations and limited target demonstrations. BEACON casts cross-domain co-training as a discrepancy-aware importance-reweighting problem, jointly learning a diffusion-based visuomotor policy and per-sample source weights that minimize an objective informed by target-domain generalization guarantees. To make best-effort adaptation practical for high-dimensional sequence policies, we develop scalable instance-level discrepancy estimators, stochastic alternating updates for policy and weights, and a multi-source extension that balances heterogeneous source domains. Across sim-to-sim, sim-to-real, and multi-source manipulation settings, BEACON improves robustness and data efficiency over target-only, fixed-ratio co-training, and feature-alignment baselines. Importantly, even without an explicit alignment objective, BEACON achieves feature alignment as an implicit result of discrepancy-aware cross-domain co-training.
title BEACON: Cross-Domain Co-Training of Generative Robot Policies via Best-Effort Adaptation
topic Robotics
url https://arxiv.org/abs/2605.08571