Conference on Cognitive Computational Neuroscience

The flexibility of intelligent behavior is fundamentally attributed to the ability to separate and assign structural information from sensory inputs. Variable binding is the atomic computation that underlies this ability. In this work, we investigate the implementation of variable binding via pointers of assemblies of neurons, which are sets of excitatory neurons that fire together. The Assembly Calculus (Papadimitriou et al PNAS 2020) is a framework that describes a set of operations to create and modify assemblies of neurons. We focus on the Reciprocal-Project operation (which performs variable binding) and study the capacity of a network in terms of the number of assemblies that can be reliably created and retrieved. We find that variable binding networks implemented through Hebbian plasticity resemble associative memories . However for networks with N neurons per brain area, the capacity of variable binding (0.01N) is an order of magnitude lower than the capacity of simple assembly creation (0.22N) through the Project operation (Fig A1). To alleviate this drop in capacity, we propose a skip connection between the input and variable assembly, which boosts the capacity to a similar order of magnitude (0.1N) as the Project operation.