Hey Timothy,

An easy way to remember is that the initial layers extract the simple features like edges, colors and as we add more layers it get sophisticated to identify larger elements or shapes until it identifies the intended object

Hope this explanation helps

Thanks!