Towards a practical measure of interference for reinforcement learning
This work addresses the challenge of understanding interference in RL systems, which is crucial for improving learning stability and efficiency, though it is incremental as it builds on existing mitigation proposals.
The paper tackles the problem of catastrophic interference in reinforcement learning by proposing a new definition and measure of interference, showing that target network frequency is a key factor and last-layer updates cause significantly higher interference, with empirical correlations to learning performance metrics like stability and sample efficiency.
Catastrophic interference is common in many network-based learning systems, and many proposals exist for mitigating it. But, before we overcome interference we must understand it better. In this work, we provide a definition of interference for control in reinforcement learning. We systematically evaluate our new measures, by assessing correlation with several measures of learning performance, including stability, sample efficiency, and online and offline control performance across a variety of learning architectures. Our new interference measure allows us to ask novel scientific questions about commonly used deep learning architectures. In particular we show that target network frequency is a dominating factor for interference, and that updates on the last layer result in significantly higher interference than updates internal to the network. This new measure can be expensive to compute; we conclude with motivation for an efficient proxy measure and empirically demonstrate it is correlated with our definition of interference.