On the Convergence of Jacobian-Free Backpropagation for Optimal Control Problems with Implicit Hamiltonians

arXiv:2602.00921v2 Announce Type: replace-cross Abstract: Optimal feedback control with implicit Hamiltonians poses a fundamental challenge for learning-based value function methods due to the absence of closed-form optimal control laws. Recent work~ite{gelphman2025end} introduced an implicit deep learning approach using Jacobian-Free Backpropagation (JFB) to address this setting, but only established sample-wise descent guarantees. In this paper, we establish convergence guarantees for JFB in the stochastic minibatch setting, showing that the resulting updates converge to stationary points of the expected optimal control objective. We further demonstrate scalability on substantially higher-dimensional problems, including multi-agent optimal consumption and swarm-based quadrotor and bicycle control. Together, our results provide both theoretical justification and empirical evidence for using JFB in high-dimensional optimal control with implicit Hamiltonians.

Related

• Accelerating Reinforcement Learning for Wind Farm Control via Expert Demonstrations
• Timescale Separation Enables Deep Reinforcement Learning Control of Rotating Detonation Engine Mode Transitions
• Live LTL Progress Tracking: Towards Task-Based Exploration
• GRAIL: Autonomous Concept Grounding for Neuro-Symbolic Reinforcement Learning
• Scalable Quantum Reinforcement Learning on NISQ Devices with Dynamic-Circuit Qubit Reuse and Grover Optimization

Source: arXiv cs.LG | 2026-04-28