Building Production Systems with Realistic Spiking Neurons

We present a novel cognitive architecture built from realistic spiking neurons that exhibits basic production system capabilities. It uses Holographic Reduced Representations to encode structured information and the Neural Engineering Framework to create detailed neuro-biologically plausible networks of spiking neurons for storing and manipulating these representations. We demonstrate this system's abilities to encode IF-THEN rules and manipulate its own representations in response to the current state. This leads to predictions about the sorts of rules that would be difficult for neurons to encode, and the maximum complexity of such rules. Our system bridges the gap between high-level cognitive architectures (including ACT-R) and modern neuroscience research, allowing details such as the numbers, types, and connections of neurons to be related to cognitive behaviour.