mlsanityData to 14 Oct 2026

Multitask LQG Control: Performance and Generalization Bounds

arXiv 2604.16730 ↗First posted 17 Apr 2026math.OCcited by 1

On the map: Deep reinforcement learning methods → RL for games and combinatorial optimization

Abstract (arXiv)

We study multitask learning for stochastic and partially observed control systems, focusing on the linear quadratic Gaussian (LQG) problem. Our goal is to learn a common stabilizing controller that generalizes across a distribution of systems and objectives. To this end, we leverage a history-dependent lifting that recasts the multitask LQG problem into an equivalent high-dimensional multitask LQR problem, allowing for the analysis of policy gradient methods. We show that learning a common lifted controller induces a heterogeneity bias which we characterize via a "bisimulation function". We establish performance and generalization guarantees that explicitly depend on such bisimulation-based heterogeneity measures. For model-free, we demonstrate that multitask learning reduces policy gradient estimation variance proportionally to the number of tasks in the training set.

Key facts. Cited by 1 paper in the map (data through 14 Oct 2026). Placed in RL for games and combinatorial optimization, one of 15 directions in Deep reinforcement learning methods. Builds on 23 of its 39 references that are in the map. First posted on arXiv 17 Apr 2026.

Builds on

23 of 39 references.

Closest ancestors · 4

Same direction, newest first

  1. 2025Policy Gradient Bounds in Multitask LQR23 Sep
  2. 2025Policy Gradient for LQR with Domain Randomization31 Mar
  3. 2024Asynchronous Heterogeneous Linear Quadratic Regulator Design13 Apr
  4. 2024Meta-Learning Linear Quadratic Regulators: A Policy Gradient MAML Approach for Model-free LQR25 Jan

Other references · 19

  1. 2020Deep Reinforcement Learning for Autonomous Driving: A Surveycited by 721
  2. 2016CVXPY: A Python-Embedded Modeling Language for Convex Optimizationcited by 641
  3. 2018Global Convergence of Policy Gradient Methods for the Linear Quadratic Regulatorcited by 250
  4. 2018Derivative-Free Methods for Policy Optimization: Guarantees for Linear Quadratic Systemscited by 67
  5. 2019Convergence and sample complexity of gradient methods for the model-free linear quadratic regulator problemcited by 46
  6. 2021Generalization Bounds for Meta-Learning via PAC-Bayes and Uniform Stabilitycited by 30
  7. 2021Analysis of the Optimization Landscape of Linear Quadratic Gaussian (LQG) Controlcited by 28
  8. 2023Model-free Learning with Heterogeneous Dynamical Systems: A Federated LQR Approachcited by 23
  9. 2021Learning over All Stabilizing Nonlinear Controllers for a Partially-Observed Linear Systemcited by 14
  10. 2021A Theoretical Understanding of Gradient Bias in Meta-Reinforcement Learningcited by 12
  11. 2023Imitation and Transfer Learning for LQG Controlcited by 8
  12. 2022Convergence and Sample Complexity of Policy Gradient Methods for Stabilizing Linear Systemscited by 6
7 more
  1. 2025Coreset-Based Task Selection for Sample-Efficient Meta-Reinforcement Learningcited by 6
  2. 2022Globally Convergent Policy Gradient Methods for Linear Quadratic Control of Partially Observed Systemscited by 5
  3. 2024Rate-Optimal Non-Asymptotics for the Quadratic Prediction Error Methodcited by 3
  4. 2024Output-feedback Synthesis Orbit Geometry: Quotient Manifolds and LQG Direct Policy Optimizationcited by 2
  5. 2025Layered Multirate Control of Constrained Linear Systemscited by 2
  6. 2026Probabilistic Performance Guarantees for Multi-Task Reinforcement Learningcited by 1
  7. 2026Generalizable Optimal Control with Transformers: One Policy Across Diverse Systemscited by 1

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2026: 1.

1 paper.

  1. 2026Why Does Agentic Safety Fail to Generalize Across Tasks?cited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2025On the Gradient Domination of the LQG Problem
  2. 2026Cost-Driven Representation Learning for Linear Quadratic Gaussian Control: Part II
  3. 2024Implicit Bias of Policy Gradient in Linear Quadratic Control: Extrapolation to Unseen Initial States
  4. 2025Policy Gradient Method for LQG Control via Input-Output-History Representation: Convergence to O(ε)-Stationary Points
  5. 2024Sample Complexity of the Linear Quadratic Regulator: A Reinforcement Learning Lens
  6. 2026Learning and Control Beyond Linearity: Towards a Non-asymptotic Theory for Bilinear Systems
  7. 2026Gradient Dominance in the Linear Quadratic Regulator: A Unified Analysis for Continuous-Time and Discrete-Time Systems
  8. 2025Comparative Study of Q-Learning for State-Feedback LQG Control with an Unknown Model