GRSN: Gated Recurrent Spiking Neurons for POMDPs and MARL
This addresses the problem of high computational resource requirements in RL for energy-constrained applications, though it is an incremental improvement by adapting SNNs to better handle temporal data.
The paper tackled the temporal mismatch between spiking neural networks (SNNs) and reinforcement learning (RL) by proposing a novel temporal alignment paradigm (TAP) that uses single-step updates and gated units to accumulate historical state information, achieving similar performance to recurrent neural networks (RNNs) with about 50% lower power consumption in solving POMDPs and multi-agent cooperation problems.
Spiking neural networks (SNNs) are widely applied in various fields due to their energy-efficient and fast-inference capabilities. Applying SNNs to reinforcement learning (RL) can significantly reduce the computational resource requirements for agents and improve the algorithm's performance under resource-constrained conditions. However, in current spiking reinforcement learning (SRL) algorithms, the simulation results of multiple time steps can only correspond to a single-step decision in RL. This is quite different from the real temporal dynamics in the brain and also fails to fully exploit the capacity of SNNs to process temporal data. In order to address this temporal mismatch issue and further take advantage of the inherent temporal dynamics of spiking neurons, we propose a novel temporal alignment paradigm (TAP) that leverages the single-step update of spiking neurons to accumulate historical state information in RL and introduces gated units to enhance the memory capacity of spiking neurons. Experimental results show that our method can solve partially observable Markov decision processes (POMDPs) and multi-agent cooperation problems with similar performance as recurrent neural networks (RNNs) but with about 50% power consumption.