Figure 1

(a) The qubit potential during application of the microwave pulses, which cause coherent transitions between states $|0⟩$ and $|1⟩$. (b) During a partial measurement the state $|1⟩$ tunnels out of the well with probability $p$. (c) The tunneling probability $p$ (solid line) is determined by the amplitude of the measurement pulse which lowers the barrier; we use sufficiently small amplitude to avoid tunneling from the state $|0⟩$ (dotted line). The maximal measurement visibility (the difference between the solid and dotted lines) is about 90%. (d) Pulse sequences (including state tomography) for the partial-collapse experiment (upper trace, as in 5) and for the quantum uncollapsing (lower trace). The effect of the partial measurement [step (i)] is undone by applying the $\pi$ pulse [step (ii)] and additional (uncollapsing) partial measurement [step (iii)] with the same strength $p$.Reuse & Permissions

Figure 2

The qubit tunneling probabilities $P_X$, $P_Y$, and $P_Z$ after the partial and tomographic ($X$, $Y$, $Z$) measurements for (a) the partial-collapse sequence, (b) the uncollapsing sequence, and (c) a “wrong” uncollapsing with $\pi$ pulse replaced by $0.9\pi$ pulse. Initial state is $(|0⟩+|1⟩)/\sqrt{2}$. The background $P_B$ is the probability of qubit tunneling before the state tomography (see text).Reuse & Permissions

Figure 3

The qubit states on the Bloch sphere, as measured by the state tomography, after the partial collapse (first row) and uncollapsing (second row). Initial state is $|1⟩$ for panels (a) and (e), $(|0⟩-i|1⟩)/\sqrt{2}$ for (b) and (f), $(|0⟩+|1⟩)/\sqrt{2}$ for (c) and (g), and $|0⟩$ for (d) and (h). The points shown on the Bloch spheres (connected by lines as a guide for the eye) correspond to varied measurement strength up to $p=0.7$. For ideal uncollapsing the states in the second row should not depend on $p$. The insets in each panel show the measured (dots) and theoretical (line) dependence of the polar angle $\theta /\pi$ on $p$.Reuse & Permissions

Figure 4

(a) The quantum process tomography matrix $\chi$ for the uncollapsing with $p=0.47$. (b) The fidelity of the quantum uncollapsing as a function of the partial measurement probability $p$ [25].Reuse & Permissions