† Corresponding author. E-mail:
Project supported by the National Key R&D Program of China (Grant No. 2017YFA0305200), the National Natural Science Foundation of China (Grant Nos. 11674234 and 11605205), the Natural Science Foundation of Chongqing, China (Grant Nos. cstc2015jcyjA00021 and cstc2018jcyjAX0656), Innovation Project of Sichuan University, China (Grant No. 2018SCUH0021), the Youth Innovation Promotion Association, Chinese Academy of Sciences (Grant No. 2015317), the Entrepreneurship and Innovation Support Program for Chongqing Overseas Returnees, China (Grant Nos. cx2017134 and cx2018040), the Fund of CAS Key Laboratory of Microscale Magnetic Resonance, and the Fund of CAS Key Laboratory of Quantum Information, China.
We introduce a modified weak value that is related to the mean value of input meter variable. With the help of the modified weak value, the validity conditions for various modified versions of weak value formalism are investigated, in terms of the dependence of the pointer shift on the mean value of the input meter. The weak value formalism, often used to represent the pointer shift, with the modified weak value is of great use in simplifying calculations and giving guidance of practical experiments whenever the mean value of the input meter variable is nonzero. The simulation in a qubit system is presented and coincident well with our theoretical result.
The concept of weak measurement was first proposed by Aharonov, Albert, and Vaidman (AAV),[1] which involves weak coupling of the system and the meter. The weak values, which could be complex and exceed the eigenvalue spectrum of observable, have been demonstrated to have important applications in exploring fundamental problems of quantum physics in Refs. [2–10]. As an important extension, the weak value amplification technique acts an essential role in precision metrology[11–13] and has been exploited to amplify minute physical quantities.[14–23,23] Recently it was also used to realize highly sensitive optical shifts in position or time domains by weak measurement with a small weak value.[24,25]
Among the practical applications of the weak measurement, it is necessary to establish an exact relationship, which was often expressed by the standard weak value formalism proposed by AAV, among the pointer shift, weak value and the coupling constant. For the standard weak value formalism (AAV formalism), which is the first order approximation in coupling strength, Jozsa gave a physical interpretation of the weak values in terms of the pointer shift.[26] The mean position shift of the pointer is divided two parts: one is proportional to the real part of the weak value and the other is proportional to the pointer spreading rate in space. Obviously, the influence of the latter is less intuitive than that of the former. It was also used to describe numerous physical phenomenons, which has resembled the optical beam propagation enhancement.[14,27] The AAV formalism related to the imaginary part of weak value was also expressed as a shift induced by varying wave function phase[28] or by covariance between the meter’s input and output variables.[12,29]
Furthermore, some other investigations beyond the AAV formalism were discussed and analyzed, such as higher-order expansions for nearly orthogonal pre- and post-selected states,[30,31] nonperturbative pre- and post-selected measurement,[29] weak measurement with the orbital angular momentum state,[32–34] the fast oscillations of large pointer shifts[35,36] and the amplification limitations of pointer shifts.[37–40]
In fact, the pointer shift is related not only to the weak value of system and the coupling constant, but to the meter’s characteristics. The previous discussions were mainly concentrated on uncertainty of meter’s input variable,[1,30,41] varying phase of wave function[12,14,26–29] and more in Refs. [32–36]. However, in some occasions the pointer shift cannot be simply expressed by the AAV formalism as shown in Refs. [2,3,18,20,23,42–48] for the dependence of the pointer shift on the central frequency. Furthermore, a series of validity conditions of the weak measurement were explored, which showed that the mean value of meter’s input variable has the same influence as the meter’s uncertainty on validity conditions of the weak measurement.[29] Hence, the effect of the mean value of meter’s input variable cannot be simply neglected because it not only affects the pointer shift but also constrains the validity of weak measurement.
Several treatments in two level systems were presented, such as the redefined system states with a rotation of the polarization of the input field according to frequency,[2,3,44,46] directly integrating the probability distributions of the final state to calculate the central frequency shifts without using the weak value formalisms.[45,48,49] Moreover, the inverse of central frequency was regarded as the interaction strength to realize phase estimation.[43] Unfortunately, a general theoretical solution has not been provided so far.
In this paper, we analyze the role of the mean value of the meter’s input variable in such a process. In order to settle this doubt, we present a modified formulas for the pointer shift, coming from the modified “standard” procedure, and a modified weak value is meanwhile introduced, related to the mean value of meter’s input variable and interaction constant, as there are also other theory about the modular value[50–52] and general approaches.[53,54] By the modified formulas, the relationship among the pointer shift, the mean value of the meter’s input variable and the validity conditions of weak measurement is clarified. Such a kind of treatment is helpful for extending the validity of weak measurement, in particular, making it capable to go on exploiting the weak value formalism with the modified weak value to precisely represent the pointer shift when the mean value of the meter’s input variable is large and necessary to be taken into account for the pointer shift. Therefore, it is of great benefit to give guidance of practical experiments and to simplify calculations whenever the mean value of input meter variable is nonzero.
The system and meter are initially prepared in the states |ψ〉 and |ψm〉, respectively. In the Von-Neumann measurement scheme, the interaction between the system and meter is described by the coupling Hamiltonian
![](cpb_29_6_064214/cpb_29_6_064214_ieqn8.gif)
![](cpb_29_6_064214/cpb_29_6_064214_ieqn9.gif)
![](cpb_29_6_064214/cpb_29_6_064214_ieqn10.gif)
In general, the pointer shift can be rewritten as[29,30]
![](cpb_29_6_064214/cpb_29_6_064214_ieqn11.gif)
![](cpb_29_6_064214/cpb_29_6_064214_ieqn12.gif)
![](cpb_29_6_064214/cpb_29_6_064214_ieqn13.gif)
![](cpb_29_6_064214/cpb_29_6_064214_ieqn14.gif)
![](cpb_29_6_064214/cpb_29_6_064214_ieqn15.gif)
Under the weak validity condition[29]
![](cpb_29_6_064214/cpb_29_6_064214_ieqn16.gif)
![](cpb_29_6_064214/cpb_29_6_064214_ieqn17.gif)
![](cpb_29_6_064214/cpb_29_6_064214_ieqn18.gif)
![](cpb_29_6_064214/cpb_29_6_064214_ieqn19.gif)
![](cpb_29_6_064214/cpb_29_6_064214_ieqn20.gif)
![](cpb_29_6_064214/cpb_29_6_064214_ieqn21.gif)
![](cpb_29_6_064214/cpb_29_6_064214_ieqn22.gif)
In the linear regime,[29] i.e. the first order approximation in γ,
![](cpb_29_6_064214/cpb_29_6_064214_ieqn23.gif)
According to the weak condition (
In order to settle this doubt, the modified weak value is introduced,
![](cpb_29_6_064214/cpb_29_6_064214_ieqn27.gif)
Based on the pointer shift form shown in Eq. (
![](cpb_29_6_064214/cpb_29_6_064214_ieqn28.gif)
By the transformations, the initial state and the postselected state of the system become ρ′ and
![](cpb_29_6_064214/cpb_29_6_064214_ieqn34.gif)
For deducing the general nonlinear formula (
By the modified procedure, the weak validity condition becomes
![](cpb_29_6_064214/cpb_29_6_064214_ieqn41.gif)
![](cpb_29_6_064214/cpb_29_6_064214_ieqn42.gif)
![](cpb_29_6_064214/cpb_29_6_064214_ieqn43.gif)
![](cpb_29_6_064214/cpb_29_6_064214_ieqn44.gif)
The linear regime condition becomes
![](cpb_29_6_064214/cpb_29_6_064214_ieqn45.gif)
The weak validity condition and the linear regime condition become
Moreover, the roles of value
We take an example to show the superiority of the weak value formalisms with the modified weak value, to represent pointer shift. Here we choose a qubit system with the operator of the system observable
![](cpb_29_6_064214/cpb_29_6_064214_ieqn55.gif)
![](cpb_29_6_064214/cpb_29_6_064214_ieqn56.gif)
![](cpb_29_6_064214/cpb_29_6_064214_ieqn57.gif)
In the following, three treatments are made to obtain the average frequency shift. First the frequency shift is calculated by direct integral ∫Ψ(ω)(ω – ω0) d ω / ∫Ψ(ω) d ω, a treatment used as comparison. The exact expression is given by
The third treatment is to represent the pointer shift with the normal weak value. The nonlinear formula (
In order to analyze the validity of the latter two treatments, the frequency shift calculated by four formulas are shown in Fig.
![]() | Fig. 2. The value ![]() |
The green dashed curve represented by nonlinear formula, with the modified weak value, shown in Eq. (
For larger values ω0, the curves for the two kinds of nonlinear formulas are shown in Fig.
Moreover, the periodicity of variation of the pointer shift with the value
By the modified “standard” procedure of the weak measurement, the extended validity of weak measurement has been verified in our simulation. Since the frequency shift would be influenced by central frequency, the former nonlinear formula (
By the modified “standard” procedure for the pointer shift, general nonlinear formulas with the modified weak value related to the mean value of the meter’s input variable and the coupling constant have been presented. The mean value
As the importance of the mean value of the meter’s input variable in weak measurements had been shown in the earlier work,[29] we give a way how to deal with it. For the role of the mean value
Our work unveils the deep connection between weak measurement and meter’s input variable, thereby making progress toward understanding and solving the problems when the mean value of meter’s input variable need be taken into account. Furthermore, we may leave a talk about if the system phase can influence the mean value of the meter’s input variable, or if the system phase can influence the meter input variable.
[1] | |
[2] | |
[3] | |
[4] | |
[5] | |
[6] | |
[7] | |
[8] | |
[9] | |
[10] | |
[11] | |
[12] | |
[13] | |
[14] | |
[15] | |
[16] | |
[17] | |
[18] | |
[19] | |
[20] | |
[21] | |
[22] | |
[23] | |
[24] | |
[25] | |
[26] | |
[27] | |
[28] | |
[29] | |
[30] | |
[31] | |
[32] | |
[33] | |
[34] | |
[35] | |
[36] | |
[37] | |
[38] | |
[39] | |
[40] | |
[41] | |
[42] | |
[43] | |
[44] | |
[45] | |
[46] | |
[47] | |
[48] | |
[49] | |
[50] | |
[51] | |
[52] | |
[53] | |
[54] |