Salvato in:
Dettagli Bibliografici
Autori principali: Lu, Shule, Wang, Yujing, Zhang, Hainan, Yang, Xiaoshan, Zheng, Hongwei, Tong, Yongxin, Xu, Changsheng, Zheng, Zhiming
Natura: Preprint
Pubblicazione: 2026
Soggetti:
Accesso online:https://arxiv.org/abs/2605.03426
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866910190942552064
author Lu, Shule
Wang, Yujing
Zhang, Hainan
Yang, Xiaoshan
Zheng, Hongwei
Tong, Yongxin
Xu, Changsheng
Zheng, Zhiming
author_facet Lu, Shule
Wang, Yujing
Zhang, Hainan
Yang, Xiaoshan
Zheng, Hongwei
Tong, Yongxin
Xu, Changsheng
Zheng, Zhiming
contents Vision-Language Models (VLMs) have broad potential in privacy-sensitive domains such as healthcare and finance, yet strict data-sharing constraints render centralized training infeasible. Federated Learning mitigates this issue by enabling decentralized training, but practical deployments face challenges due to client heterogeneity in computational resources, application requirements, and model architectures. Under extreme model and data heterogeneity, replacing parameter aggregation with preference-based collaboration offers a more suitable interface, as it eliminates the need for direct parameter or data exchange. Motivated by this, we propose MoR, a federated alignment framework that combines GRPO with Mixture-of-Rewards for heterogeneous VLMs. In MoR, each client locally trains a reward model from local preference annotations, capturing specific evaluation signals without exposing raw data. To combine these heterogeneous supervision signals, MoR introduces a Mixture-of-Rewards mechanism with learned routing, which adaptively fuses client reward models according to the input and alignment objective. The server then optimizes a base VLM using GRPO with a KL penalty to a reference model, enabling preference alignment without requiring client models to share architectures or parameters. Experiments on diverse public vision-language benchmarks demonstrate that MoR consistently outperforms federated alignment baselines in generalization and cross-client adaptability. Our approach provides a scalable solution for privacy-preserving alignment of heterogeneous VLMs under federated settings.
format Preprint
id arxiv_https___arxiv_org_abs_2605_03426
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Replacing Parameters with Preferences: Federated Alignment of Heterogeneous Vision-Language Models
Lu, Shule
Wang, Yujing
Zhang, Hainan
Yang, Xiaoshan
Zheng, Hongwei
Tong, Yongxin
Xu, Changsheng
Zheng, Zhiming
Artificial Intelligence
Vision-Language Models (VLMs) have broad potential in privacy-sensitive domains such as healthcare and finance, yet strict data-sharing constraints render centralized training infeasible. Federated Learning mitigates this issue by enabling decentralized training, but practical deployments face challenges due to client heterogeneity in computational resources, application requirements, and model architectures. Under extreme model and data heterogeneity, replacing parameter aggregation with preference-based collaboration offers a more suitable interface, as it eliminates the need for direct parameter or data exchange. Motivated by this, we propose MoR, a federated alignment framework that combines GRPO with Mixture-of-Rewards for heterogeneous VLMs. In MoR, each client locally trains a reward model from local preference annotations, capturing specific evaluation signals without exposing raw data. To combine these heterogeneous supervision signals, MoR introduces a Mixture-of-Rewards mechanism with learned routing, which adaptively fuses client reward models according to the input and alignment objective. The server then optimizes a base VLM using GRPO with a KL penalty to a reference model, enabling preference alignment without requiring client models to share architectures or parameters. Experiments on diverse public vision-language benchmarks demonstrate that MoR consistently outperforms federated alignment baselines in generalization and cross-client adaptability. Our approach provides a scalable solution for privacy-preserving alignment of heterogeneous VLMs under federated settings.
title Replacing Parameters with Preferences: Federated Alignment of Heterogeneous Vision-Language Models
topic Artificial Intelligence
url https://arxiv.org/abs/2605.03426