Tag
This paper introduces Vector-Symbolic Policy Gradient (VSPG), a novel method that uses vector symbolic architecture for discrete-action policy gradients in reinforcement learning, achieving competitive performance with robust degradation under noise for edge systems.