Short-Term Plasticity Neurons Learning to Learn and Forget

TL;DR

This paper introduces the STP Neuron, a recurrent neural unit leveraging short-term plasticity that learns to adapt, forget, and outperform existing models across various tasks, with potential energy efficiency benefits.

Contribution

The paper presents the STP Neuron, a novel recurrent unit utilizing short-term plasticity, trained via backpropagation through time, demonstrating superior performance and efficiency over existing models.

Findings

STPN outperforms RNNs, LSTMs, and other models in multiple tasks.

STPN enables learning to learn and forget in the short term.

STPN reduces energy consumption in neuromorphic circuits.

Abstract

Short-term plasticity (STP) is a mechanism that stores decaying memories in synapses of the cerebral cortex. In computing practice, STP has been used, but mostly in the niche of spiking neurons, even though theory predicts that it is the optimal solution to certain dynamic tasks. Here we present a new type of recurrent neural unit, the STP Neuron (STPN), which indeed turns out strikingly powerful. Its key mechanism is that synapses have a state, propagated through time by a self-recurrent connection-within-the-synapse. This formulation enables training the plasticity with backpropagation through time, resulting in a form of learning to learn and forget in the short term. The STPN outperforms all tested alternatives, i.e. RNNs, LSTMs, other models with fast weights, and differentiable plasticity. We confirm this in both supervised and reinforcement learning (RL), and in tasks such as Associative Retrieval, Maze Exploration, Atari video games, and MuJoCo robotics. Moreover, we calculate that, in neuromorphic or biological circuits, the STPN minimizes energy consumption across models, as it depresses individual synapses dynamically. Based on these, biological STP may have been a strong evolutionary attractor that maximizes both efficiency and computational power. The STPN now brings these neuromorphic advantages also to a broad spectrum of machine learning practice. Code is available at https://github.com/NeuromorphicComputing/stpn