Figure 4
Learning curves for two solutions and qutrit problems. (a) We depict the learning curve for the A, B, and C networks as a function of the number of episodes. Notice that the curves for the B and C networks are identical. (b) We depict the learning curve for the multitask controlled unitary operation acting on two qutrits as a function of the number of episodes. Here, \(|{\rm{in}}\rangle =|0\rangle \,|{\rm{out}}\rangle =|2\rangle \), and the solution is given by \(|{c}_{2}\rangle =|1\rangle \), where the control states coincide with the basis of the qutrit space.
