Mr. Kiernan Broda-Milian's research on a novel perturbation attack targeting reinforcement learning agents.

This digital item includes the full source code and accompanying paper, demonstrating how carefully crafted input perturbations can disrupt the policy behavior of trained agents in adversarial environments.