Exploring postselection-induced quantum phenomena with time-bidirectional state formalism Evgeniy O. Kiktenko1 2 3 4

2025-04-27 0 0 2.13MB 15 页 10玖币
侵权投诉
Exploring postselection-induced quantum phenomena
with time-bidirectional state formalism
Evgeniy O. Kiktenko1, 2, 3, 4
1Department of Mathematical Methods for Quantum Technologies,
Steklov Mathematical Institute of Russian Academy of Sciences, Moscow 119991, Russia
2Russian Quantum Center, Skolkovo, Moscow 121205, Russia
3Geoelectromagnetic Research Center, Schmidt Institute of Physics of the Earth,
Russian Academy of Sciences, Troitsk 108840, Russia
4National University of Science and Technology “MISIS”, Moscow 119049, Russia
(Dated: March 24, 2023)
Here we present the time-bidirectional state formalism (TBSF) unifying in a general manner
the standard quantum mechanical formalism with no postselection and the time-symmetrized two-
state (density) vector formalism, which deals with postselected states. In the proposed approach,
a quantum particle’s state, called a time-bidirectional state, is equivalent to a joined state of two
particles propagating in opposite time directions. For a general time-bidirectional state, we derive
outcome probabilities of generalized measurements, as well as mean and weak values of Hermitian
observables. We also show how the obtained expressions reduce to known ones in the special cases of
no postselection and generalized two-state (density) vectors. Then we develop tomography protocols
based on mutually unbiased bases and a symmetric informationally complete positive operator-
valued measure, allowing experimental reconstruction of an unknown single qubit time-bidirectional
state. Finally, we employ the developed techniques for tracking of a qubit’s time-reversal journey in
a quantum teleportation protocol realized with a cloud-accessible noisy superconducting quantum
processor. The obtained results justify an existence of a postselection-induced qubit’s proper time-
arrow, which is different from the time-arrow of a classical observer, and demonstrate capabilities
of the TBSF for exploring quantum phenomena brought forth by a postselection in the presence of
noise.
I. INTRODUCTION
The standard quantum formalism is commonly used
for calculating a probability distribution of measurement
outcomes, given a complete characterization of prepara-
tion and evolution of a measured quantum system. This
consideration with respect to a preselected initial state
contains an implicit time asymmetry related to the con-
cept of ‘collapsing’, or ‘reduction’, of system’s state in the
measurement process [1,2]. Combining the preselection
with a postselection, i.e. the consideration of a partic-
ular outcome of the measurement, removes this asym-
metry and gives rise to the two-state vector formalism
(TSVF) [2,3]. Within the TSVF, the quantum particle’s
state is described by a pair of vectors (|ψi,hφ|), where
|ψi, determined by the preselection, can be considered
evolving forward in time, while hφ|, determined by the
postselection, can be considered as evolving back from
the future to the past (see an example of an optical ex-
periment with pre- and postselection in Fig. 1).
From a practical point of view, one of the most impor-
tant concepts appearing in the framework of postselec-
tion and the TSVF is weak values of observables [5,6].
Despite some criticism (see e.g. [79]), weak values and
related techniques for a weak value amplification [1013]
appear to be extremely useful in the context of quan-
tum metrology [1423] (for a review, see [4]). Moreover,
e.kiktenko@rqc.ru
|VD|
Birefringent crystal
Single photon
source
Polarizer
(vertical)
Polarizer
(diagonal)
Detection
system
click!
Figure 1. An example of an optical experimental setup
dealing with both pre- and postselected states. Here pho-
tons pass through two polarizers, the first of which filters out
vertically polarized states |Viand the second one filters out
diagonally polarized states |Di(| hV|Hi| = 21/2). By in-
troducing a two-state vector (|Vi,hD|), the TSVF provides
a complete description of photons’ polarization, given their
detection after the second polarizer (the postselection condi-
tion). In particular, the two-state vector allows describing a
transformation of photons’ spatial degree of freedom due to
a coupling with polarization realized by a birefringent crys-
tal (see a comprehensive discussion in Ref. [4]).
taking postselection into account also plays an important
role in studying complexity theory [24], quantum contex-
tuality [2529], fundamentals of quantum physics [3040],
design of quantum computing algorithms [41], and quan-
tum communication protocols [4246].
The postselection with respect to entangled states
gives rise to time-reversal phenomena, including an ap-
pearance of closed timelike curves (CTC), considered
both theoretically [4756] and experimentally [48,53].
arXiv:2210.01583v2 [quant-ph] 23 Mar 2023
2
The basic idea behind these phenomena is that Bell state
preparation and Bell state measurement can be consid-
ered as a kind of “time mirrors” reflecting a quantum
state’s propagation in time. Note that this interpreta-
tion perfectly agrees with experimental results on delayed
entanglement swapping [57,58].
Originally, the TSVF was formulated with respect to a
pair of pure states [2,3]. An important extension comes
with introducing an ancillary particle and performing
postselection with respect to an entangled state. This
creates an entanglement between forward and backward
evolving states of a two-state vector, and yields the con-
cept of a generalized two-state vector [3,59]. Study-
ing statistical ensembles of generalized two-state vectors
bring forth a notion of two-state density vectors [60],
which can be considered as a manifestation of density ma-
trices in the framework of the TSVF. Another approach
of introducing mixedness into the TSVF is presented in
Ref. [37], where the case of forward and backward evolv-
ing states described with density matrices is considered.
The present work is devoted to a further development
of effective description of quantum states in the presence
of postselection and pursues the following two main goals.
The first goal is closing the gap between the previous
approaches for describing mixed, or randomized, quan-
tum states in the presence of the postselection and the
standard quantum formalism without any postselection
at all. This goal is achieved by extending the two-state
density vector formalism [60] with a more general time-
bidirectional state formalism (TBSF), where the posts-
election is performed with respect to an arbitrary posi-
tive operator-valued measure (POVM) effect (see Fig. 2).
Within the the developed formalism, a state of a parti-
cle is described by a bipartite, generally mixed, state,
called a time-bidirectional state, which is equivalent to a
joint state of two particles propagating in opposite time-
directions. We show that in the absence of a postse-
lection, i.e. identity postselection effect, the backward
evolving part appears to be in the maximally mixed state,
while the forward evolving one coincides with a density
matrix from the standard formalism. An important fea-
ture of the TBSF is its ability to account for any kind of
decoherence noise, affecting both pre- and postselection.
The second goal is developing practical schemes for to-
mography of both pre- and postselected states. In the
current work, we focus on the case of a single qubit that
can be easily generalized to a multiqubit one. Compared
to a high level recipe for making a complete set of Kraus
operators, sufficient for reconstructing an unknown pre-
and postselected state, presented in Ref. [60], here we
obtain explicit circuits for running tomography proto-
cols on an arbitrary quantum processor. For this pur-
pose, we borrow two basic approaches for single-qubit
tomography: the one based on mutually unbiased bases
(MUBs) corresponding to measuring three components
of a Bloch vector, and the second based on a symmetric
informationally complete POVM (SIC-POVM) allowing
reconstruction of unknown state with a measurement of
ρpre A
Q
ρprobe PUμ
pass|fail
record /
discard
μ
μ
AliceBob
μ
postselection result
(pass / fail)
Time
POVM
{E(μ )}μ
POVM with
Epost
Figure 2. General scheme of a postselection experiment giving
rise to a concept of a time-bidirectional state. First, Alice
prepares particles Q and A in some arbitrary state and sends
Q to Bob. On his side, Bob applies a unitary operation U
to Q and P, then performs an arbitrary measurement on P,
and finally returns Q back to Alice. Then Alice makes a joint
measurement on A and Q, described by a POVM containing
a particular effect Epost . If the outcome given by Epost is
realized, then Alice tells Bob to keep his measurement result
µ, otherwise µis discarded.
a single type.
To demonstrate capabilities of the TBSF and devel-
oped tomography techniques, we consider a well-known
time-reversal phenomenon appearing in a quantum tele-
portation protocol [48,50,53]. Namely, we track propa-
gation of a qubit state, initially prepared by Alice, (i) for-
ward in time to the moment of a Bell measurement on her
qubit and a qubit from a pre-shared Bell pair, then (ii)
back in time on Alice’ qubit from the Bell pair to the mo-
ment of the Bell pair birth, and then (iii) forward in time
on Bob’s particle from the Bell pair. For this purpose we
use a seven-qubit cloud-accessible noisy superconducting
quantum processor provided by IBM. Although, experi-
ments on the observation of a postselection-induced time-
travel in quantum teleportation were considered previ-
ously [48,53], to the best of our knowledge, this is the
first time where it is demonstrated, using the developed
formalism, how the state, prepared by Alice, propagates
back in time on Alice’s physical qubit taken from the
pre-shared Bell pair. As already mentioned, an impor-
tant advantage of the developed TBSF, compared, e.g.,
to the time-reversal formalism suggested in Ref. [48], is
that this formalism allows considering decoherence ef-
fects. In particular, we observe evidences of irreversible
corruption of the Alice’s quantum state during propaga-
tion along its proper postselection-induced time-arrow.
The rest of the paper is organized as follows. In Sec. II,
we introduce the concept of time-bidirectional states,
provide some illustrative examples, and derive their main
mathematical properties. In Sec. III, we discuss descrip-
tion of measurements made on a time-bidirectional state
with a focus on von Neuman measurement of Hermitian
observables and measurements of weak values. In Sec. IV,
we develop tomography protocols for experimental recon-
structing of an unknown single-qubit time-bidirectional
state. In Sec. V, we apply the developed formalism
and tomography techniques for observing a time-reversal
3
journey of a qubit’s state in a quantum teleportation pro-
tocol. We conclude and provide an outlook in Sec. VI.
II. INTRODUCING TIME-BIDIRECTIONAL
STATES
A. General postselection experiment
Let us consider a postselection experiment shown in
Fig. 2. The experiment is realized by two parties, named
Alice and Bob, that are able to communicate with quan-
tum particles and classical messages. At the start of the
experiment, Alice prepares two particles, Q and A, in an
arbitrary joint mixed state
ρpre =ρpre
ii0;mm0|iiQhi0|⊗|miAhm0|.(1)
Here |niXwith integer labels ndenote computational
basis states of particle X and ρpre
ii0;mm0are density ma-
trix elements providing standard conditions ρpre 0,
Trρpre = 1. Note that here and hereafter we apply the
Einstein summation convention and omit explicit sum-
mation with respect to repeated indices. After its prepa-
ration, Q is given to Bob, while A remains with Alice.
On his side, Bob takes an additional particle P, pre-
pared in some mixed state
ρprobe =ρprobe
kk0|kiPhk0|(2)
(ρprobe 0, Trρprobe = 1), and lets P and Q evolve
according to a unitary evolution operator
U=Uk;i
l;j|liPhk|⊗|jiQhi|.(3)
Then Bob makes a measurement on P described by
a POVM {E(µ0)}µ0, satisfying standard conditions
E(µ0)0, Pµ0E(µ0) = 1, where 1is the identity ma-
trix. Here we consider outcome labels µ0belonging to an
arbitrary finite set. Bob keeps the obtained measurement
outcome µ, corresponding to the realized effect E(µ), and
returns Q back to Alice.
On her side, Alice makes a joint measurement on Q
and A, described by another POVM, whose collection of
effects includes a particular effect
Epost =Ejj0;mm0
post |j0iQhj|⊗|m0iAhm|(4)
(0 Epost 1). If an outcome of Alice’s measurement
corresponds to Epost, then we say that the postselection
passed and set a special flag variable ps := 1; postselec-
tion failed and ps := 0 otherwise. Alice transmits the
result of the postselection, i.e. single bit ps, to Bob, who
keeps his measurement outcome µif postselection has
passed, or discards µotherwise. We note that the only
constraints on particular time moments, when the de-
scribed operations take place, correspond to the general
ordering: Preparations of ρpre and ρprobe are in the past
light cone of U, while Alice’s and Bob’s measurements
are in the future light cone of Uand in the past light
cone of the final decision on µ.
The main object of our study is a conditional proba-
bility distribution of Bob’s measurement outcomes, given
the passing postselection on Alice’s side (here Mdenotes
a random variable of Bob’s outcome). According to ax-
iomatics of quantum mechanics, the probability of the
joint event of M=µand the postselection passing is
given by
Pr[M=µps = 1]
=ρpre
ii0;kk0ρprobe
mm0Ui;m
l;jUi0;m0
l0;j0E(µ)ll0Ejj0;kk0
post ,(5)
where the overbar denotes the complex conjugate. The
probability of the postselection passing takes the form
Pr[ps = 1] = ρpre
˜
i˜
i0;˜
k˜
k0ρprobe
˜mm0U˜
i; ˜m
˜
l;˜
jU˜
i0; ˜m0
˜
l;˜
j0E˜
j˜
j0;˜
k˜
k0
post Ppost.
(6)
Putting these expressions in Bayes’ rule leads to
Pr[M=µ|ps = 1] = Pr[M=µps = 1]
Ppost
=ρpre
ii0;kk0ρprobe
mm0Ui;m
l;jUi0;m0
l0;j0E(µ)ll0Ejj0;kk0
post
ρpre
˜
i˜
i0;˜
k˜
k0ρprobe
˜mm0U˜
i; ˜m
˜
l;˜
jU˜
i0; ˜m0
˜
l;˜
j0E˜
j˜
j0;˜
k˜
k0
post P(µ),(7)
given Ppost >0, and P(µ) = 0 otherwise. It appears
that Eq. (7) is much easier to follow in the form of a
tensor network, shown in Fig. 3(a). We note that if Zis
a tensor, then by Zwe denote a tensor obtained from Z
by taking complex conjugation of all element. It differs
from the Hermitian conjugate of Z, which we denote as
Z.
Within Eq. (7) we can separate two mathematical
structures, which are related to Alice’s and Bob’s actions
correspondingly. The pre- and postselection, which are
performed by Alice, are described by a tensor
ηjj0
ii0:= ρpre
ii0;mm0Ejj0;mm0
post ,(8)
while the indirect Bob’s measurement is described by a
collection of tensors
K(µ)ii0
jj0:= ρprobe
kk0Uk;i
l;jUk0;i0
l0;j0E(µ)ll0.(9)
We then can rewrite Eqs. (7) and (6) in a compact form
P(µ) = K(µ)ii0
mm0ηmm0
ii0
K˜
i˜
i
˜m˜m0η˜m˜m0
˜
i˜
i0K(µ)η
Kη,
Ppost =Kη,
(10)
where K:= PµK(µ) and , in line with Ref. [60], de-
notes contraction with respect to proper indices [see also
Fig. 3(a)].
In what follows, we refer to η,K(µ), and Kas a time-
bidirectional state, operation outcome µtensor, and op-
eration tensor correspondingly. The time-bidirectional
摘要:

Exploringpostselection-inducedquantumphenomenawithtime-bidirectionalstateformalismEvgeniyO.Kiktenko1,2,3,41DepartmentofMathematicalMethodsforQuantumTechnologies,SteklovMathematicalInstituteofRussianAcademyofSciences,Moscow119991,Russia2RussianQuantumCenter,Skolkovo,Moscow121205,Russia3Geoelectromagn...

展开>> 收起<<
Exploring postselection-induced quantum phenomena with time-bidirectional state formalism Evgeniy O. Kiktenko1 2 3 4.pdf

共15页,预览3页

还剩页未读, 继续阅读

声明:本站为文档C2C交易模式,即用户上传的文档直接被用户下载,本站只是中间服务平台,本站所有文档下载所得的收益归上传人(含作者)所有。玖贝云文库仅提供信息存储空间,仅对用户上传内容的表现方式做保护处理,对上载内容本身不做任何修改或编辑。若文档所含内容侵犯了您的版权或隐私,请立即通知玖贝云文库,我们立即给予删除!
分类:图书资源 价格:10玖币 属性:15 页 大小:2.13MB 格式:PDF 时间:2025-04-27

开通VIP享超值会员特权

  • 多端同步记录
  • 高速下载文档
  • 免费文档工具
  • 分享文档赚钱
  • 每日登录抽奖
  • 优质衍生服务
/ 15
客服
关注