July, 2021
Neural Re-Rresentation is a computational artwork that explores enabling RL agents to represent and communicate their environment to one another. To do this in the Reinforcement Learning (RL) environment, I enabled agents to communicate through a visual plane to one another, which was also their only representation of their environment. This relation questioning how collectives of RL agents are able to represent space and work together, but through a way that humans can also see and perceive. This work in action forms beautiful abstract patterns and revolving gradients which are the RL agents playfully communicating to one another, movements and dancing with one another.
This work alongside others were presented at 4S in Cholula 2022, in both the making and doing exhibition as well as a panel called "Future imagining and the cultivation of possibility", organised by Rachel Horst, and Esteban Morales.
Check out the: