FGRPO: Federated GRPO with Adaptive Aggregation on Non-IID Data

• URL: http://arxiv.org/abs/2606.03094v1
• Date: Tue, 02 Jun 2026 03:32:32 GMT
• Title: FGRPO: Federated GRPO with Adaptive Aggregation on Non-IID Data
• Abstract summary: Group relative policy optimization (GRPO) offers superior scalability by eliminating the critic network.<n>We introduce federated GRPO, a framework designed to decentralize the fine-tuning of reasoning models across heterogeneous data owners.
• Score: 20.603813803213768
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Recent advances in language models have established reinforcement learning as the primary paradigm for eliciting self-correction and long-chain reasoning. While group relative policy optimization (GRPO) offers superior scalability by eliminating the critic network, deploying it on a central infrastructure entails collecting a large volume of data from distributed owners, which poses significant privacy risks. To address these concerns, we introduce federated GRPO (FGRPO), a framework designed to decentralize the fine-tuning of reasoning models across heterogeneous data owners. To effectively mitigate the instability caused by divergent reward scales across heterogeneous tasks, FGRPO incorporates an adaptive aggregation mechanism based on relative performance gain. By characterizing each client's improvement relative to its personalized historical baseline, the framework dynamically prioritizes effective learning trajectories regardless of local task difficulty. FGRPO ensures robust convergence on non-IID data while preserving data privacy.

Related papers

• Replacing Parameters with Preferences: Federated Alignment of Heterogeneous Vision-Language Models [63.70401095689976]
  Vision-Language Models (VLMs) have broad potential in privacy-sensitive domains such as healthcare and finance.<n>We propose MoR, a federated alignment framework that combines GRPO with Mixture-of-Rewards for heterogeneousVLMs.<n>MoR consistently outperforms federated alignment baselines in generalization and cross-client adaptability.
  arXiv  Detail & Related papers  (2026-05-05T07:02:50Z)
• FedGRPO: Privately Optimizing Foundation Models with Group-Relative Rewards from Domain Client [21.08829811371245]
  Existing methods based on model level or representation level knowledge transfer either require expensive local training or incur high communication costs.<n>We reformulate this problem as a reinforcement learning style evaluation process and propose FedGRPO.<n>FedGRPO achieves superior downstream accuracy and communication efficiency compared to conventional FedFMs baselines.
  arXiv  Detail & Related papers  (2026-02-12T14:45:56Z)
• Replacing Parameters with Preferences: Federated Alignment of Heterogeneous Vision-Language Models [63.70401095689976]
  We argue that replacing parameters with preferences represents a more scalable and privacy-preserving future.<n>We propose MoR, a federated alignment framework based on GRPO with Mixture-of-Rewards for heterogeneous VLMs.<n>MoR consistently outperforms federated alignment baselines in generalization, robustness, and cross-client adaptability.
  arXiv  Detail & Related papers  (2026-01-31T03:11:51Z)
• FedMOA: Federated GRPO for Personalized Reasoning LLMs under Heterogeneous Rewards [16.516565435444445]
  GRPO's critic-free architecture enables feasible on-device training, yet transitioning to a federated setting introduces systemic challenges.<n>We propose FedMOA, a federated GRPO framework for multi-objective alignment under heterogeneous rewards.
  arXiv  Detail & Related papers  (2026-01-31T02:00:17Z)
• MAESTRO: Meta-learning Adaptive Estimation of Scalarization Trade-offs for Reward Optimization [56.074760766965085]
  Group-Relative Policy Optimization has emerged as an efficient paradigm for aligning Large Language Models (LLMs)<n>We propose MAESTRO, which treats reward scalarization as a dynamic latent policy, leveraging the model's terminal hidden states as a semantic bottleneck.<n>We formulate this as a contextual bandit problem within a bi-level optimization framework, where a lightweight Conductor network co-evolves with the policy by utilizing group-relative advantages as a meta-reward signal.
  arXiv  Detail & Related papers  (2026-01-12T05:02:48Z)
• Generalized and Personalized Federated Learning with Foundation Models via Orthogonal Transformations [4.008780119020479]
  Federated Learning aims to train models across decentralized clients or devices holding local data without the need for centralized data collection.<n>We introduce FedOT, a novel approach that leverages black-box foundation models.<n>FedOT mitigates gradient conflicts across diverse clients, preserves semantic integrity, and achieves robust performance even in the presence of substantial data.
  arXiv  Detail & Related papers  (2025-05-26T12:18:24Z)
• DISCO Balances the Scales: Adaptive Domain- and Difficulty-Aware Reinforcement Learning on Imbalanced Data [65.09939942413651]
  We propose a principled extension to GRPO that addresses inter-group imbalance with two key innovations.<n> Domain-aware reward scaling counteracts frequency bias by reweighting optimization based on domain prevalence.<n>Difficulty-aware reward scaling leverages prompt-level self-consistency to identify and prioritize uncertain prompts that offer greater learning value.
  arXiv  Detail & Related papers  (2025-05-21T03:43:29Z)
• DisCO: Reinforcing Large Reasoning Models with Discriminative Constrained Optimization [50.91849555841057]
  Group Relative Policy Optimization is a reinforcement learning method for large reasoning models (LRMs)<n>We introduce a new Discriminative Constrained Optimization framework for reinforcing LRMs, grounded in the principle of discriminative learning.<n>DisCO significantly outperforms GRPO and its improved variants such as DAPO, achieving average gains of 7% over GRPO and 6% over DAPO.
  arXiv  Detail & Related papers  (2025-05-18T11:08:32Z)
• Privacy-Preserving Federated Embedding Learning for Localized Retrieval-Augmented Generation [60.81109086640437]
  We propose a novel framework called Federated Retrieval-Augmented Generation (FedE4RAG)<n>FedE4RAG facilitates collaborative training of client-side RAG retrieval models.<n>We apply homomorphic encryption within federated learning to safeguard model parameters.
  arXiv  Detail & Related papers  (2025-04-27T04:26:02Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.