SmartConnections We built an RL agent, specifically trained using Deep Q-learning, to efficiently solve the NYT Connections Puzzle. We used Word2Vec to parametize our values and generate our embeddings