Exploring the Promise and Limits of Real-Time Recurrent Learning

arXiv:2305.19044v2 Announce Type: replace
Abstract: Real-time recurrent learning (RTRL) for sequence-processing recurrent neural networks (RNNs) offers certain conceptual advantages over backpropagation through time (BPTT). RTRL requires neither caching past activations nor truncating context, and enables online learning. However, RTRL's time and space complexity make it impractical. To overcome this problem, most recent work on RTRL focuses on approximation theories, while experiments are often limited to diagnostic settings. Here we explore the practical promise of RTRL in more realistic settings. We …

abstract advantages arxiv backpropagation bptt caching complexity context cs.lg networks neural networks online learning processing real-time recurrent neural networks space through type