memoriae is a sculptural work built from the data of 500 simulated agents. Using C++ and the Allolib library (built by UCSB's Allosphere Research Group for performance in the AlloSphere), I took inspiration from the Boids flocking algorithm, developed by Craig Reynolds in 1986, to drive the interaction between agents. The simulation is also evolutionary, where agents have lifespans that is increased upon consuming food particles in their environment. Each agent's individual fitness is assessed, determined by a set of rules that quantify viability or desirability of an agent's flocking parameters (which start off as random values, but through reproduction, are inherited by the offspring to eventually lead to a stabilized system where flocking patterns are homogenous). I wanted to physically illustrate the path of an agent's motion with geometry, so I recorded a minute of run-time of my program (~3600 frames of data) with 500 agents. Bringing in this data to the Rhinoceros and Grasshopper environments, I was able to visualize one agent's pattern of motion as it is affected by interactions with other agents, resulting in a finalized geometry that serves as a memory or snapshot of these interactions over time. This work is currently in progress, as I would like to create one geometry for each of the 500 agents and physically print each form.