Multiplicative couplings facilitate rapid learning and information gating in recurrent neural networks
The brain consists of reciprocal connectivity and loops between recurrent neural networks (RNNs) and feedforward neural networks (FNNs). However, how their interactions facilitate learning remains unknown. Here we propose a multiplicative RNN-FNN coupling mechanism and report remarkable computational strengths in learning. The multiplicative interaction imposes a Hebbian-weight amplification onto synaptic-neuronal coupling, enabling context-dependent gating and rapid switching. We demonstrate that multiplicative coupling-driven synaptic plasticity achieves 2-100 folds of speed improvement in supervised, reinforcement and unsupervised learning settings, boosting memory capacity, model robustness and generalization of RNNs. We further demonstrate the efficacy and biological plausibility of multiplicative gating in modeling multiregional circuits, including a prefrontal cortex-mediodorsal thalamus network for context-dependent decision making, a cortico-thalamic-cortical network for working memory and attention, and an entorhinal cortex-hippocampus network for visuospatial navigation and sequence replay. Take together, our results offer insights into multi-plasticity, attractor dynamics and computation of recurrent neural circuits and profound neuroscience-inspired applications.