mlsanityData to 14 Oct 2026

Gradient Inversion in Federated Reinforcement Learning

arXiv 2512.00303 ↗First posted 29 Nov 2025cs.LG cs.AIcited by 1

On the map: Deep reinforcement learning methods → RL for energy, networks and operations

Abstract (arXiv)

Federated reinforcement learning (FRL) enables distributed learning of optimal policies while preserving local data privacy through gradient sharing.However, FRL faces the risk of data privacy leaks, where attackers exploit shared gradients to reconstruct local training data.Compared to traditional supervised federated learning, successful reconstruction in FRL requires the generated data not only to match the shared gradients but also to align with real transition dynamics of the environment (i.e., aligning with the real data transition distribution).To address this issue, we propose a novel attack method called Regularization Gradient Inversion Attack (RGIA), which enforces prior-knowledge-based regularization on states, rewards, and transition dynamics during the optimization process to ensure that the reconstructed data remain close to the true transition distribution.Theoretically, we prove that the prior-knowledge-based regularization term narrows the solution space from a broad set containing spurious solutions to a constrained subset that satisfies both gradient matching and true transition dynamics.Extensive experiments on control tasks and autonomous driving tasks demonstrate that RGIA can effectively constrain reconstructed data transition distributions and thus successfully reconstruct local private data.

Key facts. Cited by 1 paper in the map (data through 14 Oct 2026). Placed in RL for energy, networks and operations, one of 15 directions in Deep reinforcement learning methods. Builds on 11 of its 27 references that are in the map. First posted on arXiv 29 Nov 2025.

Builds on

11 of 27 references.

References · 11

  1. 2022Generative Adversarial Networkscited by 5,627
  2. 2016Asynchronous Methods for Deep Reinforcement Learningcited by 2,258
  3. 2021Offline Reinforcement Learning with Implicit Q-Learningcited by 1,293
  4. 2018Off-Policy Deep Reinforcement Learning without Explorationcited by 1,241
  5. 2019Deep Leakage from Gradientscited by 1,021
  6. 2020iDLG: Improved Deep Leakage from Gradientscited by 382
  7. 2021See through Gradients: Image Batch Recovery via GradInversioncited by 326
  8. 2021Evaluating Gradient Inversion Attacks and Defenses in Federated Learningcited by 196
  9. 2023GIFD: A Generative Gradient Inversion Method with Feature Domain Optimizationcited by 45
  10. 2021Multi-Task Federated Reinforcement Learning with Adversariescited by 13
  11. 2023Local Environment Poisoning Attacks on Federated Reinforcement Learningcited by 6

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2026: 1.

1 paper.

  1. 2026Temporal Gradient Inversion for Private Trajectory Reconstruction in Embodied Reinforcement Learningcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2025GUIDE: Enhancing Gradient Inversion Attacks in Federated Learning with Denoising Models
  2. 2024Gradient Inversion of Federated Diffusion Models
  3. 2022Gradient Obfuscation Gives a False Sense of Security in Federated Learning
  4. 2025Enhanced Privacy Leakage from Noise-Perturbed Gradients via Gradient-Guided Conditional Diffusion Models
  5. 2024Federated Learning Nodes Can Reconstruct Peers' Image Data
  6. 2024Global Convergence Guarantees for Federated Policy Gradient Methods with Adversaries
  7. 2025KL-regularization Itself is Differentially Private in Bandits and RLHF
  8. 2025POLAR: Policy-based Layerwise Reinforcement Learning Method for Stealthy Backdoor Attacks in Federated Learning