Enabling Realtime Reinforcement Learning At Scale With Staggered Asynchronous Inference

Abstract

Realtime environments change even as agents perform action inference and learning, thus requiring high interaction frequencies to effectively minimize regret. However, recent advances in machine learning involve larger neural networks with longer inference times, raising questions about their applicability in realtime systems where reaction time is crucial. We present an analysis of lower bounds on regret in realtime reinforcement learning (RL) environments to show that minimizing long-term regret is generally impossible within the typical sequential interaction and learning paradigm, but often becomes possible when sufficient asynchronous compute is available. We propose novel algorithms for staggering asynchronous inference processes to ensure that actions are taken at consistent time intervals, and demonstrate that use of models with high action inference times is only constrained by the environment's effective stochasticity over the inference horizon, and not by action frequency. O

Stats

• citations0
• S2 citations—
• github stars0
• HF likes0
• heat score0.00
• arxiv keyriemer2024enabling

Related papers

• The Fallacy Of Minimizing Cumulative Regret In The Sequential Task Setting (2024)0.00
• Revisiting State Augmentation Methods For Reinforcement Learning With Stochastic Delays (2021)10.35
• ACERAC: Efficient Reinforcement Learning In Fine Time Discretization (2021)4.52
• Demystifying Reinforcement Learning In Time-varying Systems (2022)0.00
• Dynamic Memory For Interpretable Sequential Optimisation (2022)0.00
• Deep Reinforcement Learning With Time-scale Invariant Memory (2024)0.00
• Reinforcement Learning With Fast Stabilization In Linear Dynamical Systems (2020)0.00
• Learning When To Act: Interval-aware Reinforcement Learning With Predictive Temporal Structure (2026)0.00