Arxiv:1711.01208V2 [Quant-Ph] 25 Apr 2018

Dynamics of a qubit while simultaneously monitoring its relaxation and dephasing

Q. Ficheux,1, 2 S. Jezouin,2 Z. Leghtas,3, 2, 4 and B. Huard1, 2, ∗ 1UniversitéLyon, ENS de Lyon, UniversitéClaude Bernard, CNRS, Laboratoire de Physique, F-69342 Lyon, France 2Laboratoire Pierre Aigrain, Département de physique de l’ENS, Ecole´ normale supérieure, PSL Research University, UniversitéParis Diderot, Sorbonne Paris Cité,Sorbonne Universités, UPMC Univ. Paris 06, CNRS, 75005 Paris, France 3Centre Automatique et Systèmes,Mines ParisTech, PSL Research University, 60 Boulevard Saint-Michel, 75272 Paris Cedex 6, France. 4QUANTIC team, INRIA de Paris, 2 Rue Simone Iff, 75012 Paris, France (Dated: April 26, 2018) Abstract Decoherence originates from the leakage of quantum information into external degrees of freedom. For a qubit the two main decoherence channels are relaxation and dephasing. Here, we report an experiment on a superconducting qubit where we retrieve part of the lost information in both of these channels. We demonstrate that raw averaging the corresponding measurement records provides a full quantum tomography of the qubit state where all three components of the effective spin-1/2 are simultaneously measured. From single realizations of the experiment, it is possible to infer the quantum trajectories followed by the qubit state conditioned on relaxation and/or dephasing channels. The incompatibility between these quantum measurements of the qubit leads to observable consequences in the statistics of quantum states. The high level of controllability of superconducting circuits enables us to explore many regimes from the Zeno effect to underdamped Rabi oscillations depending on the relative strengths of driving, dephasing and relaxation.

Introduction fore extending the previously explored case of two in- Decoherence can be understood as the result of mea- compatible measurement outcomes [10, 12] to the case of surement of a system by its environment. For a qubit, three spin directions. By varying the drive amplitudes at the two main sources of decoherence are relaxation by the cavity and qubit transition frequencies, we are able to spontaneous emission and dephasing that can be mod- reach a variety of regimes corresponding to different con- eled by unmonitored readout of coupled quantum sys- figurations for Ω/Γ1 and Γd/Γ1, where Ω is the Rabi fre- tems (Fig. 1a). What becomes of the qubit state if, quency, Γ1 the fixed relaxation rate and Γd the dephasing instead of disregarding the information leaking to the rate. This work hence provides a textbook experimen- environment, we continuously monitor both decoherence tal demonstration of quantum measurement backaction channels? Owing to measurement backaction, the knowl- on a qubit with incompatible and simultaneous measure- edge of the measurement record then leads to a stochas- ments. tic quantum trajectory of the qubit state for each single Results realization of an experiment [1–3]. Recently, diffusive Description of the experiment. Two parallel de- quantum trajectories were observed following the contin- tection setups operate via spatially separated measure- uous homodyne or heterodyne measurements of either a ment lines (see Fig. 1b). The fluorescence heterodyne dephasing channel [4–9] or a relaxation channel [10, 11]. detection setup enables the measurement of both quadratures u(t) and v(t) of the spontaneously emitted field out Here we report an experiment in which we have si- of a 3D transmon qubit [13]. The complex amplitude of multaneously monitored the spontaneous emission of a

arXiv:1711.01208v2 [quant-ph] 25 Apr 2018 the emitted field is on average proportional to the expec- superconducting qubit by heterodyne measurement (re- tation of the qubit lowering operator σ = (σ − iσ )/2 laxation channel) and the transmitted field through a − x y so that u and v are on average proportional to the ex- dispersively coupled cavity by homodyne measurement pectations of σ and σ . Here, σ = |gi he| + |ei hg|, (dephasing channel). We demonstrate that the average x y x σ = i |gi he| − i |ei hg| and σ = |ei he| − |gi hg| are the outcomes of these two non-projective measurements are y z qubit Pauli operators, where |gi and |ei are the ground the three coordinates x, y and z of the Bloch vector. It and excited states. For a single realization, the measure- is remarkable that a full quantum tomography can be ment outcomes read [10] obtained at any time by simply raw averaging measure- p ment outcomes of many realizations of a single experi- u(t)dt = ηf Γ1/2x(t)dt + dWu(t) ment despite the incompatibility of the three observables p , (1) v(t)dt = ηf Γ1/2y(t)dt + dWv(t) that characterize a qubit state. For single realizations the −1 resulting quantum trajectories show signatures of the in- where Γ1 = (15 µs) is the qubit relaxation rate, compatibility between the measurement channels, there- ηf = 0.14 is the total fluorescence measurement efficiency, 2

excited state (the qubit and cavity are in the dispersive regime and χcq = 5.1 MHz as explained in Supplemen- tary Note 2). The phase of the measured quadrature in the homodyne measurement can then be chosen in such a way that [8, 17] p w(t)dt = 2ηdΓdz(t)dt + dWw(t). (2)

Importantly, the measurement induced dephasing rate Γd can be tuned arbitrarily as it is proportional to the drive power at fd. Similarly to the notations above, ηd = 0.34 is the total dispersive measurement efficiency, z(t) = Tr(σzρt) is the last of the three Bloch coordinates (see Fig. 1a) and Ww is another independent Wiener process. Full tomography by direct averaging. As can be seen from Eqs. (1,2), taking a raw average of the FIG. 1: Measurement setup and quantum trajectory result- outcomes (u, v, w) on a large number of realizations of ing from its outputs. a Bloch vector representation of a the experiment directly leads to the Bloch coordinates qubit whose state is described by a density matrix ρt = (x, y, z) of the qubit. In Fig. 2, we show the direct aver- (1 + x(t)σx + y(t)σy + z(t)σz) /2. A quantum trajectory ρt is represented as a blue line. The qubit decoherence can be aging of the three outcomes in two configurations of the modeled as originating from a relaxation channel at a rate input drives: one in the regime of underdamped Rabi os- Γ1 and a dispersive measurement channel at a rate Γd. b A cillations (Fig. 2a) and another in the regime of strong superconducting qubit in a cavity is driven by two microwave dispersive measurement rate, the so-called Zeno regime signals at the weakly coupled input. The one at qubit fre- (Fig. 2b). The raw averaging of (u, v, w), once rescaled by quency fq = 5.353 GHz (orange) induces Rabi oscillations the prefactors in Eqs. (1,2), agrees well with the average of the qubit at frequency Ω. The one at cavity frequency evolution of the qubit, as predicted by the solution of the f = 7.761 GHz (purple) leads to a dispersive measurement d master equation (Eq. 3 below without the last stochastic of the qubit state along σz. A diplexer at the strongly coupled output port separates the outgoing signals depending on term). We thus demonstrate that performing a dispersive their frequency. The radiation at fq that is spontaneously measurement and a measurement of fluorescence reveals emitted by the qubit is processed by a Josephson Paramet- information on all three components of a spin-1/2. Such a ric Converter (JPC) [14, 15] so that a following heterodyne direct full tomography cannot be done by measuring two measurement reveals the two quadratures u(t) and v(t) of the records only [10, 12]. Note however that it is possible to fluorescence field [10, 16, 17]. The transmitted signal at fd perform an indirect tomography using a small number of is processed by a doubly pumped Josephson Parametric Am- plifier (JPA) [4, 18] with a pump phase such that a following records and maximum likelihood estimation [19]. A com- homodyne measurement reveals the quadrature w(t) of the parison between our technique and the usual technique field at fd. c Measurement records u (blue), v (red) and w using a qubit rotation followed by a projective measure- (yellow) as a function of time for one realization of the exper- ment is discussed in Supplementary Note 3. iment. These records feed the stochastic master equation (3), Experiments of Figs. 2a and 2b differ by the relative which leads to the trajectory in a rate of the dispersive readout Γd compared to the Rabi frequency Ω. For weak measurement rate Γd, Γ1 < Ω, the Rabi oscillations are underdamped, while they are x(t) = Tr(σxρt) and y(t) = Tr(σyρt) are the qubit Bloch overdamped when Γd Ω, Γ1 owing to the fact that the coordinates corresponding to the density matrix ρt (see Zeno effect prevents any unitary evolution such as Rabi Fig. 1a) and Wu and Wv are two independent stochas- oscillations. For a single realization of the experiment tic Wiener processes describing the measurement noise, though, the trajectory of the qubit state that one can which includes the zero point fluctuations, and such that infer from the measurement records u(t), v(t) and w(t) dW 2 = dt and dW is zero on average. Experimen- can strongly differ from this average behavior. tally, the measurement takes a non infinitesimal time Single quantum trajectories. In order to determine dt = 100 ns, which we chose smaller than the inverse this quantum trajectory, one can use the formalism of the measurement rates and compatible with the detection stochastic master equation [17]. The density matrix at bandwidth (see Supplementary Note 4). time t + dt can be decomposed into ρt+dt = ρt + dρt, Similarly, the dispersive detection setup (see Fig. 1b) where enables the measurement of a single quadrature w(t) of Ω X X √ the transmitted field at frequency fd = fr −χcq/2, which dρ = i[ σ , ρ ]dt+ D (ρ )dt+ η M (ρ )dW (t), t 2 y t k t k k t k is between the cavity resonance frequencies fr and fr−χcq k k respectively corresponding to a qubit in the ground and (3) 3

FIG. 2: Direct averaging of the three measurement records. FIG. 3: Tomographic validation of the quantum trajectories. a Dots: Rescaled average of the measurement records a,b,c Correlations between the coordinates (xtraj, ytraj, ztraj) p p u˜(t) =u ¯(t)/ ηf Γ1/2,v ˜(t) =v ¯(t)/ ηf Γ1/2 andw ˜(t) = of the trajectories after 19.8 µs of evolution and an indepen- √ 6 w¯(t)/ 2ηdΓd for 1.5 10 realizations of an experiment where dent tomography on the dataset corresponding to the experi- the qubit starts in |gi at time 0 and is driven so that it rotates ment of Fig. 2a. Each panel reprensents the average value of −1 at a Rabi frequency Ω/2π = (2 µs) around σy and endures the tomography results for the subset of trajectories ending −1 a measurement induced dephasing rate Γd = (5 µs) . Lines: up less than 0.01 away from a given value of xtraj (a), ytraj Calculated coordinates of the Bloch vector x(t), y(t) and z(t) (b) or ztraj (c). The error bars are given by the standard de- from the master equation (Eq. 3 with ηi = 0). b Same figure viation of the tomography results divided by the square root in the Zeno regime with a drive such that Ω/2π = (16 µs)−1 of the number of trajectories in the subset (out of a total −1 and Γd = (0.9 µs) . number of 1.5 million trajectories per panel). The agreement between the tomography and the coordinates of the trajectories demonstrates the validity of the quantum trajectories. with the four Lindblad superoperator (k ∈ {u, v, w, ϕ}) d Bloch sphere representation of 3 quantum trajectories that end up with 0.74 < xtraj < 0.76 (red dashed line) after 19.8 µs 1 1 corresponding to one bin of the histogram in a. D (ρ ) = L ρ L† − ρ L† L − L† L ρ , (4) k t k t k 2 t k k 2 k k t and the measurement backaction superoperator realizations of the experiment for which the trajectory † † predicts a given value x(T ) = xtraj at a time T . If the Mk(ρt) = Lkρt + ρtLk − Tr(Lkρt + ρtLk)ρt. (5) trajectories are valid, then a strong measurement of σx at In these expressions, the jump operators corresponding time T should give x on average on this post-selected p traj to heterodyning fluorescence are Lu = Γ1/2σ− and ensemble of realizations (Fig. 3a). We have checked for p Lv = i Γ1/2σ− and the jump operator correspond- any value of xtraj, ytraj and ztraj (Fig. 3), and for 30 rep- ing to homodyning the dispersive measurement is Lw = resentative configurations of drives that the trajectories p p Γd/2σz. A fourth jump operator Lϕ = Γϕ/2σz predict the strong measurement results (Supplementary corresponds to the unread (ηϕ = 0) pure dephasing Note 1). In fact, we found that the agreement is verified of the qubit, so that the total decoherence rate Γ2 = for efficiencies ηf and ηd within a confidence interval of Γ1 −1 ±0.02 for any of the 30 configurations. 2 + Γϕ + Γd can be tuned from Γ2 = (11.2 µs) to higher arbitrary values depending on the power of the Evolution of the distribution of states. Any mea- drive at frequency fd. Interestingly, the two fluorescence surement record is a stochastic process and the corre- measurement records u and v exert a different backac- sponding quantum trajectories follow a random walk in tion but act identically on average (same Lindblad oper- the Bloch sphere with a state dependent diffusion con- ators). The additional dispersive measurement that we stant. The inherent backaction of a quantum measure- introduced compared to Ref. [10] thus leads to a very ment is thus better discussed by representing distribu- different dynamics. tions of states at a given time [4, 5, 8, 10–12, 20] or Using this formalism it is possible to reconstruct the distributions of trajectories for a given duration [7, 21– quantum trajectory of the qubit state in time from any 24]. Figure 4 gives a different perspective to the Rabi set of measurement records (see Fig. 1a,c in the case oscillation of Fig. 2a by representing the distributions of −1 −1 where Ω/2π = (5.2 µs) and Γd = (0.9 µs) ). The the qubit states conditioned on the three measurement validity of the reconstructed quantum trajectories can records u(t), v(t) and w(t) for 1.5 million realizations be tested independently by post-selecting an ensemble of of the experiment. In the Supplementary Note 1, one 4

# trajectories 1 10 100 1000

FIG. 4: Evolution of the distribution of quantum states. a,b,c Colored dots: Each frame represents the marginal distribution, in the x − y (a), x − z (Fig. b) and y − z (Fig. c) planes of the Bloch sphere, of the states of the qubit at a given time τ for 1.5 million realizations of the experiment, in the same experimental conditions as Fig. 2a. Each state (x, y, z) is reconstructed from the measurement records {u(t), v(t), w(t)} from time t between 0 to τ using Eq. (3). Time τ is increasing from 0.3 µs to 19 µs from left to right as indicated at the bottom of the figure. For each figure, the surrounding black circles represent the pure states of the plane (e.g. z = 0 for a). Solid lines: average projection of all 1.5 millions of quantum trajectories {x(t), y(t), z(t)} for 0.2 µs < t < τ. can find movies of the distributions of 1.5 millions ex- Interplay between detectors. Interestingly, while perimental realizations for each configuration of the Rabi the average trajectory stays in the x − z plane with frequency Ω and the dephasing rate Γd for a set of 30 dif- hσyi = 0, the backaction of the fluorescence measure- ferent experimentally realized configurations. Evidently, ment leads to a nonzero spread in the y direction of the the Rabi drive term −Ωσy/2 still provides an overall an- Bloch sphere. This competition between the backaction gular velocity in the x − z plane of the Bloch sphere. of relaxation (fluorescence measurement) and dephasing However, the measurement backaction is such that some (dispersive measurement) measurements can be better trajectories are delayed while others are advanced com- observed when decoherence dominates the dynamics. In pared to the average evolution. As time increases the Fig. 5, we show the distributions of qubit states at a long spread in the qubit states grows as a result of the cumu- time τ = 6.5 µs after which the distribution is close to its lated effect of the stochastic measurement backaction at steady state while the qubit is both Rabi driven and dis- each time step. persively measured at a strong measurement rate. The The effect of decoherence under a strong Rabi drive trajectories are determined using three sets of measure- corresponds to an average loss of purity, defined as ment records: dispersive only {w(t)}, fluorescence only Tr(ρ2) = (1 + x2 + y2 + z2)/2 and it can be seen as a {u(t), v(t)} or both. As in Fig. 2b, the Zeno effect then decreasing distance of the mean trajectory from the cen- leads to the dampening of the Rabi oscillations and the ter of the Bloch sphere when time increases (solid line). average trajectory (solid line) quickly reaches its steady When the dispersive measurement (dephasing channel) state. is measured in presence of the Rabi drive around σy, the In contrast, a trajectory corresponding to a single real- corresponding distribution of states tends to be uniform ization of the experiment where the dispersive measure- in the x − z plane at long times (right panel in Fig. 4b), ment w(t) is recorded is found to consist in a series of which is similar to what is obtained by simultaneously stochastic jumps between two areas of the Bloch sphere measuring σx and σz in an effectively undriven qubit [12]. that are close to the two eigenstate of the σZ measure- The experiment thus illustrates the fact that the average ment operator. In the distribution of states, this leads loss of purity corresponds to the statistical uncertainty to two areas with high probability of occupation near on the quantum state when the decoherence channel is the poles of the Bloch sphere. These areas can be in- unread. terpreted as zones frozen by the Zeno effect. The rest of 5

dispersive measurement (Figs. 5d-f), where at long times the qubit states spans a small ball in the Bloch sphere. Therefore, the com- a b c bined action of Rabi drive and fluorescence measurement backaction leads to a uniform spread of the qubit state close to the most entropic state 1/2 at the center of the sphere. As expected, the quantum states that are conditioned on all measurement records {u(t), v(t), w(t)} are less entropic than with a single measurement. This can fluorescence measurement be seen in Figs. 5g-i where the spread of the distributions is larger than for the cases of single measurements d e f Figs. 5a-f. A clear asymmetry appears in the spread of the marginal distribution in the x − y plane of Fig. 5g between positive and negative values of x. This asymmetry originates from the fact that the fluorescence measurement is linked to the jump operator σ− for which dispersive and fluorescence measurement |gi is the single pointer state. Indeed the measurement backaction is null when the qubit state is close to |gi g h i (Mu(|gihg|) = Mv(|gihg|) = 0) while it is strongest when the qubit state is close to |ei. Since the Rabi drive cor- relates the ground state to positive x (red zone shifted to the right of the south pole in Fig. 5h) and the excited state to negative x, the spread in y is smaller for positive x than for negative x. This asymmetry highlights the profound difference between measuring both quadra- # trajectories tures of fluorescence and measuring σx and σy simulta- 1 10 100 1000 neously using dynamical states as in Ref. [12, 26]. While both methods lead to the same result on average, their FIG. 5: Impact of the type of detector on the distribution of quantum states. a,b,c Marginal distribution in the x, y (a), backaction differs. The latter corresponds to quantum x − z (b) and y − z (c) planes of the Bloch sphere of the non-demolition measurements, while fluorescence does qubit states ρτ corresponding to 1.5 millions of measurement not. In the end, the asymmetry in the distributions of records at the cavity frequency only {w(t)} from time t be- Figs. 5g,i results from the incompatibility between a dis- tween 0 and τ = 6.5 µs. The information about {u(t), v(t)} persive measurement with no backaction on |ei and a is here discarded (ηf = 0). All panels in the figure correspond −1 −1 fluorescence measurement with maximal backaction on to the Zeno regime (Ω/2π = (5.2 µs) Γd = (0.9 µs) ) |ei. As in Fig. 4, the boundary of the Bloch sphere is represented as a black circle and the average quantum trajectory as a In conclusion, we have shown quantum trajectories of solid line. d,e,f Case where the states are conditioned on a superconducting qubit reconstructed from three mea- fluorescence records {u(t), v(t)} instead while discarding the surements originating from the simultaneous monitoring information on {w(t)} (ηd = 0). g,h,i Case where the states of its decoherence channels. It looks promising to test are conditioned on both fluorescence and dispersive measure- statistical properties of quantum trajectories [27, 28], ment records {u(t), v(t), w(t)}. fluctuation relations in quantum thermodynamics [29– 35], quantum smoothing protocols [20, 36–41], and to perform parameter estimation [42, 43]. the Bloch sphere is still occupied with a lower probability Data availability The experiment was carried out for 30 exper- (Fig. 5b) because of the finite time it takes for the jump imental configurations with Ω/2π ranging from 0 to (2 µs)−1 and −1 −1 to occur from one pole to the next under strong disper- Γd ranging from (30 µs) to (300 ns) . All the experimental results can be visualized in a small animated application available sive measurement rate [25]. Note how the ensemble of online at trajectories can go from uniform for weak measurement http://www.physinfo.fr/publications/Ficheux1710.html. rates (Fig. 4b rightest panel) to localized at the poles for The measurement can be chosen to take into account the mea- strong measurement rates (Fig. 5b). surement records of the dispersive measurement only, the fluorescence measurement only or both. The movies are also available to As can be seen from Figs. 5a,c, the dispersive measure- download at ment alone does not provide any backaction towards the https://doi.org/10.6084/m9.figshare.6127958.v1. y direction of the Bloch sphere so that the qubit states All raw data used in this study are available from the corre- keeps a zero σ component during its evolution. This sponding authors upon reasonable request. y Author contributions Q.F. and S.J. contributed equally to is in stark contrast with the trajectories corresponding this work. Q.F., S.J. and B.H. designed research and performed to measurement records {u(t), v(t)} of the fluorescence research; S.J. and Q.F. analyzed data; ZL contributed to the ex- 6 perimental setup; All authors wrote the paper. [14] N. Bergeal, F. Schackert, M. Metcalfe, R. Vijay, V. E. Acknowledgements We thank Philippe Campagne-Ibarcq, Manucharyan, L. Frunzio, D. E. Prober, R. J. Schoelkopf, Michel Devoret, Andrew Jordan, RaphaëlLescanne, Mazyar Mir- S. M. Girvin and M. H. Devoret. Phase-preserving am- rahimi, Klaus Mølmer, Pierre Rouchon, Alain Sarlette, Irfan Sid- plification near the quantum limit with a Josephson ring diqi and Pierre Six for fruitful interactions. Nanofabrication has modulator. Nature 465, 64 (2010). been made within the consortium Salle Blanche Paris Centre. This [15] N. Roch, E. Flurin, F. Nguyen, P. Morfin, P. Campagne- work was supported by the EMERGENCES grant QUMOTEL of Ibarcq, M. H. Devoret and B. Huard. Widely Tun- Ville de Paris. able, Nondegenerate Three-Wave Mixing Microwave De- Competing Interests The Authors declare no Competing Fi- vice Operating near the Quantum Limit. Physical Review nancial or Non-Financial Interests. Letters 108, 147701 (2012). [16] P. Campagne-Ibarcq, L. Bretheau, E. Flurin, A. Auffèves, F. Mallet and B. Huard. Observing Interferences between Past and Future Quantum States in Resonance Fluores- cence. Phys. Rev. Lett. 112, 180402 (2014). ∗ Electronic address: [email protected] [17] P. Campagne-Ibarcq. Measurement back action and feed- [1] H. Carmichael. An Open Systems Approach to Quantum back in superconducting circuits. Ph.D. thesis, Ecole Nor- Optics (Springer Berlin Heidelberg, 1993). male Supérieure (ENS) (2015). [2] A. Barchielli and M. Gregoratti. Quantum Trajecto- [18] A. Kamal, A. Marblestone and M. Devoret. Signal-tories and Measurements in Continuous Time, vol. 782 pump back action and self-oscillation in double-pump (Springer-Verlag Berlin Heidelberg, 2009). Josephson parametric amplifier. Physical Review B 79, [3] H. M. Wiseman and G. J. Milburn. Quantum Measure- 184301 (2009). ment and Control (Cambridge University Press, 2009). [19] R. L. Cook, C. A. Riofr´ıoand I. H. Deutsch. Single-shot [4] K. W. Murch, S. J. Weber, C. Macklin and I. Siddiqi. Ob- quantum state estimation via a continuous measurement serving single quantum trajectories of a superconducting in the strong backaction regime. Physical Review A 90, quantum bit. Nature 502, 211–214 (2013). 032113 (2014). [5] M. Hatridge, S. Shankar, Mirrahimi, F. Schackert, K. [20] D. Tan, S. Weber, I. Siddiqi, K. Mølmer and K. Murch. Geerlings, T. Brecht, K. Sliwa, B. Abdo, L. Frunzio, Prediction and Retrodiction for a Continuously Moni- Girvin, Schoelkopf and M. Devoret. Quantum Back- tored Superconducting Qubit. Physical Review Letters Action of an Individual Variable-Strength Measurement. 114, 090403 (2015). Science 339, 178 (2013). [21] A. N. Jordan, A. Chantasri, P. Rouchon and B. Huard. [6] G. de Lange, D. Ristè, M. Tiggelman, C. Eichler, L. Anatomy of Fluorescence: Quantum trajectory statis- Tornberg, G. Johansson, a. Wallraff, R. Schouten and L. tics from continuously measuring spontaneous emission. DiCarlo. Reversing Quantum Trajectories with Analog Quantum Studies: Mathematics and Foundations 3, 237 Feedback. Physical Review Letters 112, 080501 (2014). (2016). [7] S. J. Weber, A. Chantasri, J. Dressel, A. N. Jordan, [22] M. Naghiloo, D. Tan, P. M. Harrington, P. Lewalle, K. W. Murch and I. Siddiqi. Mapping the optimal route A. N. Jordan and K. W. Murch. Quantum caustics in between two quantum states. Nature 511, 570–3 (2014). resonance-fluorescence trajectories. Physical Review A [8] S. J. Weber, K. W. Murch, M. E. Kimchi-Schwartz, N. 96, 053807 (2017). Roch and I. Siddiqi. Quantum trajectories of supercon- [23] A. Chantasri, M. E. Kimchi-Schwartz, N. Roch, I. Sid- ducting qubits. Comptes Rendus Physique 17, 766–777 diqi and A. N. Jordan. Quantum Trajectories and Their (2016). Statistics for Remotely Entangled Quantum Bits. Phys- [9] A. Chantasri, M. E. Kimchi-Schwartz, N. Roch, I. Sid- ical Review X 6, 041052 (2016). diqi and A. N. Jordan. Quantum Trajectories and Their [24] P. Lewalle, A. Chantasri and A. N. Jordan. Prediction Statistics for Remotely Entangled Quantum Bits. Phys- and characterization of multiple extremal paths in contin- ical Review X 6, 041052 (2016). uously monitored qubits. Physical Review A 95, 042126 [10] P. Campagne-Ibarcq, P. Six, L. Bretheau, A. Sarlette, M. (2017). Mirrahimi, P. Rouchon and B. Huard. Observing Quan- [25] Y. Choi and A. N. Jordan. Operational approach to in- tum State Diffusion by Heterodyne Detection of Fluores- directly measuring the tunneling time. Physical Review cence. Phys. Rev. X 6, 11002 (2016). A 88, 052128 (2013). [11] M. Naghiloo, N. Foroozani, D. Tan, A. Jadbabaie and [26] U. Vool, S. Shankar, S. O. Mundhada, N. Ofek, A. Narla, K. W. Murch. Mapping quantum state dynamics in K. Sliwa, E. Zalys-Geller, Y. Liu, L. Frunzio, R. J. spontaneous emission. Nature Communications 7, 11527 Schoelkopf, S. M. Girvin and M. H. Devoret. Contin- (2016). uous quantum nondemolition measurement of the trans- [12] S. Hacohen-Gourgy, L. S. Martin, E. Flurin, V. V. Ra- verse component of a qubit. Physical Review Letters 117, masesh, K. B. Whaley and I. Siddiqi. Dynamics of simul- 133601 (2016). taneously measured non-commuting observables. Nature [27] A. Franquet and Y. V. Nazarov. Probability distributions 538, 491 (2016). of continuous measurement results for conditioned quan- [13] H. Paik, D. I. Schuster, L. S. Bishop, G. Kirchmair, tum evolution. Physical Review B 95, 085427 (2017). G. Catelani, a. P. Sears, B. R. Johnson, M. J. Reagor, [28] A. Chantasri, J. Atalaya, S. Hacohen-Gourgy, L. S. Mar- L. Frunzio, L. I. Glazman, S. M. Girvin, M. H. De- tin, I. Siddiqi and A. N. Jordan. Simultaneous continuous voret and R. J. Schoelkopf. Observation of High Coher- measurement of noncommuting observables: Quantum ence in Josephson Junction Qubits Measured in a Three- state correlations. Physical Review A 97, 012118 (2018). Dimensional Circuit QED Architecture. Physical Review [29] Z. Gong, Y. Ashida and M. Ueda. Quantum-trajectory Letters 107, 240501 (2011). thermodynamics with discrete feedback control. Physical 7

Review A 94, 012107 (2016). [37] M. Tsang. Optimal waveform estimation for classical and [30] C. Elouard, D. A. Herrera-Mart´ı, M. Clusel and A. quantum systems via time-symmetric smoothing. Phys- Auffèves. The role of quantum measurement in stochastic ical Review A 80, 033840 (2009). thermodynamics. npj Quantum Information 3, 9 (2017). [38] S. Gammelmark, B. Julsgaard and K. Mølmer. Past [31] J. J. Alonso, E. Lutz and A. Romito. Thermodynamics Quantum States of a Monitored System. Physical Re- of Weakly Measured Quantum Systems. Physical Review view Letters 111, 160401 (2013). Letters 116, 080403 (2016). [39] I. Guevara and H. Wiseman. Quantum State Smoothing. [32] P. G. Di Stefano, J. J. Alonso, E. Lutz, G. Falci and M. Physical Review Letters 115, 180407 (2015). Paternostro. Non-equilibrium thermodynamics of contin- [40] D. Tan, M. Naghiloo, K. Mølmer and K. W. Murch. uously measured quantum systems: a circuit-QED imple- Quantum smoothing for classical mixtures (2016). mentation. arxiv:1704.00574 (2017). [41] D. Tan, N. Foroozani, M. Naghiloo, A. H. Kiilerich, K. [33] S. Deffner, J. P. Paz and W. H. Zurek. Quantum work Mølmer and K. W. Murch. Homodyne monitoring of and the thermodynamic cost of quantum measurements. postselected decay. Physical Review A 96, 022104 (2017). Physical Review E 94, 010103 (2016). [42] P. Six, P. Campagne-Ibarcq, L. Bretheau, B. Huard and [34] M. Naghiloo, D. Tan, P. M. Harrington, J. J. Alonso, P. Rouchon. Parameter estimation from measurements E. Lutz, A. Romito and K. W. Murch. Thermody- along quantum trajectories. In Decision and Control namics along individual trajectories of a quantum bit. (CDC), 2015 IEEE 54th Annual Conference on, 7742– arxiv:1703.05885 (2017). 7748 (IEEE, 2015). [35] F. Liu. Heat and work in Markovian quantum master [43] A. H. Kiilerich and K. Mølmer. Bayesian parameter es- equations: concepts, fluctuation theorems, and compu- timation by continuous homodyne detection. Physical tations. Progress in Physics 38, 1–62 (2018). Review A 94, 032103 (2016). [36] M. Tsang. Time-Symmetric Quantum Theory of Smooth- ing. Physical Review Letters 102, 250403 (2009).