Assuming I use NEAT to evolve a neural network after 1000 generations, how can I then use the same winning network to kickstart the next round of generations so that the algorit