BEACON: Cross-Domain Co-Training of Generative Robot Policies via Best-Effort Adaptation

¹Brown University · ²Harvard University

Abstract

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.

BEACON overview: cross-domain policy co-training via best-effort adaptation. — **BEACON.** Given a large amount of (multi-)source demonstrations and a small amount of target demonstrations, BEACON jointly learns a diffusion-based visuomotor policy and a set of per-sample weights that directly minimizes a theoretically grounded discrepancy-based generalization bound on the target domain.

Supplementary Video

Citation

@article{zhang2026beacon,
  title   = {BEACON: Cross-Domain Co-Training of Generative Robot Policies via Best-Effort Adaptation},
  author  = {Zhang, Antong and Qi, Han and Yang, Heng},
  year    = {2026}
  eprint = {2605.08571}, 
  archivePrefix = {arXiv},
  primaryClass  = {cs.RO},
  url           = {https://arxiv.org/abs/2605.08571},
}