We discuss here some simple examples where we introduce what is called 'memory'about previous steps, or what is normally called momentum gradient descent. For the mathematical details, see whiteboad notes from lecture on September 8, 2025.