Comments on this demonstration run of a GA
Population size way too small
Good features combine through crossover
New features added through mutation
Initial population is randomly generated and diverse
Population becomes more similar as the run proceeds
The output of the GA run is the best member in the population when the run ends