Transient Non-Stationarity and Generalisation in Deep Reinforcement Learning
This addresses generalization issues in deep RL for researchers and practitioners, but it is incremental as it builds on existing methods to mitigate a known bottleneck.
The paper tackles the problem of transient non-stationarity in deep reinforcement learning, which can harm generalization, and proposes Iterated Relearning (ITER) to improve performance on benchmarks like ProcGen and Multiroom.
Non-stationarity can arise in Reinforcement Learning (RL) even in stationary environments. For example, most RL algorithms collect new data throughout training, using a non-stationary behaviour policy. Due to the transience of this non-stationarity, it is often not explicitly addressed in deep RL and a single neural network is continually updated. However, we find evidence that neural networks exhibit a memory effect where these transient non-stationarities can permanently impact the latent representation and adversely affect generalisation performance. Consequently, to improve generalisation of deep RL agents, we propose Iterated Relearning (ITER). ITER augments standard RL training by repeated knowledge transfer of the current policy into a freshly initialised network, which thereby experiences less non-stationarity during training. Experimentally, we show that ITER improves performance on the challenging generalisation benchmarks ProcGen and Multiroom.