This project implements a Deep Q-Network (DQN) agent trained in a custom Gymnasium environment called HomeFoodEnvStatic. The agent learns to navigate a grid-based world to reach a desired food goal while avoiding distractions ("hell states") using reinforcement learning.