Abstract: We present a Reinforcement Learning (RL) framework for automated inverse design of resonant structures. To accelerate the training of the RL-agent, the framework utilizes a surrogate neural ...