Abstract. Spector et al. have shown [1],[2],[3] that genetic programming can be used to evolve quantum circuits. In this paper, we present new results in this field, introducing probabilistic and deterministic quantum circuits that have not been previously published. We compare our techniques with those of Spector et al, and point out some differences in perspective between our two approaches. Finally, we show how, by using sets of functions rather than precise quantum states as fitness cases, our basic technique can be extended to evolve true quantum algorithms.
Paul Massey, John A. Clark, Susan Stepney