Retrodiction of measurement outcomes on a single quantum system reveals entanglement with its environment
Julien Pinske, Klaus Mølmer · 2025
Read Retrodiction of measurement outcomes on a single quantum system reveals entanglement with its environment by Julien Pinske, Klaus Mølmer (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2504-16170.txt.
Introduction
Julien Pinske and Klaus Molmer wrote this 2025 paper on reading a quantum system's past to learn about its ties. A standard math tool predicts what a measurement on a quantum system will give. It cannot tell plain chance in setup from a hidden link to the surroundings. The authors use retrodiction, reasoning backward with both earlier and later knowledge.
They form backward-looking chances for results on a single system. How sharply those past results can be pinned down turns out to witness a link with the environment. The certainty measure also gauges how strong that link is. The work gives experimenters a way to probe entanglement without measuring the environment itself, using records from one small system alone.
Catalog record
| CREATOR | Julien Pinske, Klaus Mølmer |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 55 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2504-16170
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Retrodiction of measurement outcomes on a single quantum system reveals entanglement with its environment Julien Pinske, Klaus Mølmer Abstract The density matrix yields probabilistic information about the outcome of measurements on a quantum system, but it does not distinguish between classical randomness in the preparation of the system and entanglement with its environment. Here, we show that retrodiction, employing both prior and posterior knowledge, gives rise to conditional probabilities for measurements on a single system, that can witness if it is part of a larger composite system. The degree of certainty with which one can retrodict the outcomes of multiple measurements on a system can witness both the existence and the quantitative nature of its entanglement with the environment. \Omega_{m} m \rho_{A} B A (a) \rho B A E \Omega_{m} m (b) Figure 1: Verifying the existence of entanglement within a composite quantum system AB . Measurements performed on subsystem A with outcomes m are associated with the application of the operator \Omega_{m} . (a) The composite system AB is prepared in the initial state \rho , and the probabilities for the outcomes m are determined by the reduced density matrix \rho_{A}=\mathrm{Tr}_{B}(\rho) , allowing no conclusions about the entanglement with B . (b) If the composite system is initially prepared in \rho and is ultimately detected as described by a measurement effect E , the retrodicted, or postselected, probabilities for the outcomes m on A can reveal entanglement, imposed by \rho and E within the composite system. I Introduction In the treatment of open quantum systems, a system A experiences loss of excitation and coherence, due to its interaction and entanglement with its environment B . The probabilities of the outcomes of measurements on the whole system AB are governed by the density matrix \rho_{AB} , while for measurements on the smaller system, they depend only on the reduced density matrix \rho_{A}=\mathrm{Tr}_{B}(\rho_{AB}) . Similarly, entangled states shared by users of a quantum network appear mixed, when local measurement outcomes are not communicated between the parties involved, and approaches to determine global properties need access to joint correlations [ 1 , 2 , 3 ] . The situation is different if we have access to both the initial density matrix \rho_{AB} and the result of a posterior joint measurement on the composite system AB . This situation is associated with the process of postselection, and for the case of pure initial and final states, separate expressions exist for the outcome of weak [ 4 ] and projective measurements [ 5 ] . In a more general setting, the probability of the outcome of any kind of measurement at any time may be conditioned on the initial density matrix, the known Hamiltonian and dissipative evolution of the system and the results of any sequence of earlier and later measurements on the system [ 6 , 7 ] . This includes, for example, the monitoring of an open quantum system, such as continuous homodyning by photodetection [ 8 ] , and the estimation of unknown classical perturbations, applied to a probed quantum system [ 9 , 10 ] . Even though we only address measurements on A , the retrodicted outcome probabilities are not captured by the reduced density matrix \rho_{A} but by the full density matrix \rho_{AB} and the results of later observations on the system AB . An illustration of this phenomenon is the remarkable fact that it is possible to pass a single spin-1/2 particle to an experimentalist, let them do an experiment and return the particle for subsequent inspection, and tell with certainty what was the outcome of the measurement of any of the three Cartesian spin components \sigma^{x} , \sigma^{y} , and \sigma^{z} [ 11 ] . For two spin components such retrodiction is simple: first provide an eigenstate of \sigma^{x} and measure \sigma^{y} of the returned particle, then the intermediate measurement of \sigma^{x} or \sigma^{y} would have to agree with the prior or posterior value, respectively. The retrodiction of three candidate measurement outcomes is possible when the spin is initially entangled with another spin, and they are finally measured to be in an entangled state [ 11 , 12 ] . Although the full density matrix \rho_{AB} , and a complete account for the postselection by joint measurements on A and B seem necessary for retrodiction of measurements on system A , we show that it is possible to capture the same information using a statistical operator acting only on the small system. This offers a picture where the entanglement between the system and its environment is witnessed by inseparable correlations between state matrices and measurement effects on the probe system. Our analysis suggests that free (distillable) entanglement [ 13 ] between A and B is necessary for the sharp retrodiction of multiple noncommuting observables. Beyond verifying entanglement, we are also interested in determining the dimension of the external (inaccessible) system. To address this issue in a systematic way, we appropriate the concept of a dimension witness for postselected states. Originally applied to Bell inequalities [ 14 , 15 , 16 ] , dimension witnesses were employed in prepare-and-measure scenarios [ 17 , 18 ] and in the analysis of experimental data [ 19 , 20 , 21 ] . Existing approaches to witnessing the dimension of the environment of a system require joint measurements to go beyond the predictions offered by the reduced density matrix. The theory developed here for postselected or retrodicted outcomes bypasses this limitation. We note that questions concerning the discernibility of different ensemble realizations of a reduced density matrix was recently discussed from a similar retrodiction perspective [ 22 ] . The article is structured as follows. Section II provides a brief review of the past quantum state formulation of retrodiction and postselection. As our first key result, we derive a reduced operator for the accessible part of the system in Sec. III . It is shown that this operator’s positivity yields a separability criteria. In Sec. IV , we apply the reduced operator to continuous-variable systems and show how its phase-space representation can be used to efficiently calculate outcome probabilities. Section V introduces dimension witnesses for past states, the second key result of this paper. In Sec. VI the theory is applied to measurements of mutually unbiased bases. Finally, Sec. VII provides a summary of the article and concluding remarks. II Past quantum states Here, we review the notion of past quantum states [ 6 ] . Let \rho be a density matrix accounting for prior knowledge about a system. A measurement (POVM [ 23 ] ) is described by a collection of operators \Omega_{m} satisfying \Omega_{m}^{\dagger}\Omega_{m}\geq 0 and \sum_{m}\Omega_{m}^{\dagger}\Omega_{m}=\mathbb{1} , where \mathbb{1} is the identity and m labels different outcomes. The probability for obtaining m is given by Born’s rule p(m)=\mathrm{Tr}(\Omega_{m}\rho\Omega^{\dagger}_{m}) and the state after the measurement reads \Omega_{m}\rho\Omega_{m}^{\dagger}/p(m) . If a second measurement, described by operators \Lambda_{n} , is performed, we iterate the above formalism and find that the joint probability of obtaining first outcome m and then outcome n , is p(m,n)=\mathrm{Tr}(\Lambda_{n}\Omega_{m}\rho\Omega^{\dagger}_{m}\Lambda^{\dagger}_{n}) . With this joint probability at hand, we can infer the probability that the early measurement yielded outcome m conditioned on any definite outcome n^{\prime} :