Exact and approximate conditions of tabletop reversibility: when is Petz recovery cost-free?

Minjeong Song, Hyukjoon Kwon, Valerio Scarani · 2025

Read Exact and approximate conditions of tabletop reversibility: when is Petz recovery cost-free? by Minjeong Song, Hyukjoon Kwon, Valerio Scarani (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-2510-26895.txt.

Introduction

Minjeong Song, Hyukjoon Kwon, and Valerio Scarani ask when undoing a noisy quantum channel can be done with the same tabletop resources as the channel itself. Channels that describe open quantum dynamics cannot be reversed fully, but the Petz recovery map gives a systematic partial recovery tied to a reference state. Normally its implementation looks very different from the channel's.

Song and her two co-authors study tabletop reversibility, where the recovery is realisable with similar resources. For the exact case, they show time-sensitive control of an ancilla system is needed. For the approximate case, they give conditions that drop that requirement. They also derive conditions for Lindbladian dynamics under a random-time collision model. The results clarify when recovery is cost-free in practice.

Catalog record

CREATORMinjeong Song, Hyukjoon Kwon, Valerio Scarani
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE97 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2510-26895

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Exact and approximate conditions of tabletop reversibility: when is Petz recovery cost-free? Minjeong Song, Hyukjoon Kwon, Valerio Scarani Abstract Channels 𝒩 \mathcal{N} that describe open quantum dynamics are inherently irreversible: it is impossible to undo their effect completely, but one can study partial recovery of the information. The Petz recovery map 𝒩 ^ γ ( P ) \hat{\mathcal{N}}_{\gamma}^{(\texttt{P})} is a systematic construction that depends only on 𝒩 \mathcal{N} and on a reference state γ \gamma , which will be recovered exactly. If the real input state was different from γ \gamma , the recovery is partial, with a guarantee of near-optimality. Generically, an implementation of the Petz recovery map would look very different from the implementation of the channel. It is natural to study under which conditions the two maps require similar or even identical resources. The noisy forward channel 𝒩 \mathcal{N} is called tabletop time-reversible for a given γ \gamma when the corresponding Petz recovery map is realizable in such a way. First, we study the exact tabletop reversibility (TTR) conditions. We show in particular that a time-sensitive control of an ancilla system is needed. Second, we present the approximate TTR conditions, which do not require such a time-sensitive control. Third, we derive Lindbladian TTR conditions under a random-time collision model. I Introduction All practical quantum information tasks confront dissipative noise from the surrounding bath over time Breuer and Petruccione 2007 . This inevitable phenomenon leads to irreversibility of noisy quantum dynamics. Nonetheless, such noisy quantum dynamics can be recovered by considering recovery channels for some quantum states. A recovery channel, known as a Petz recovery map , has particularly proven useful in many research areas. Quantum error correction is one of its prominent applications since a Petz recovery map for a code space is shown to be the optimal recovery when the code is perfectly correctable Barnum and Knill 2002 . Moreover, it performs nearly optimal compared to existing quantum error correcting codes together with recovery algorithms when the code is not too far from perfect codes Barnum and Knill 2002 ; Ng and Mandayam 2010 ; Chen et al. 2020 ; Cree and Sorce 2022 ; Zheng et al. 2024 ; Biswas et al. 2025 ; Li et al. 2025 . It is a natural figure of merit for a recovery what states can be exactly recovered, and if approximate, how close the recovered state is from the initial state before noise. The former has been answered in the original paper by Petz Petz 1986 ; Petz 1988 (see also Ref. Petz 2003 for a less technical proof). It was shown that a Petz recovery map successfully recovers states when they satisfy some entropic conditions, i.e., when the relative entropy between the state to recover and a so-called reference state remains unchanged under the noisy dynamics Hayden et al. 2004 . An approximation of this condition concerns a question as to how the changes in the relative entropy affect the distance between the initial state and the recovered state. A Petz recovery map and its variants have then enabled an information theoretic formulation of recoverability in terms of relative entropies Wilde 2015 ; Sutter et al. 2016a ; Junge et al. 2018 . They also made appearance in a plethora of studies on recoverability in the special case concerning a tripartite system in terms of the conditional mutual information Fawzi and Renner 2015 ; Sutter et al. 2016b ; Kim 2021 ; Hu and Zou 2024 ; Kim et al. 2024 . Namely, if a conditional mutual information between two subsystems conditioning the other is zero, a Petz recovery map can recover a global state from a local state of the two subsystems only and the global state is called a quantum Markov chain Hayden et al. 2004 . A Petz recovery map manifests in many other quantum foundation. To name a few, it has been adopted as a crucial tool for formulating quantum fluctuation theorems Åberg 2018 ; Kwon and Kim 2019 ; Aw et al. 2021 ; Buscemi and Scarani 2021 and has appeared as one of the promising candidates of quantum Bayesian retrodiction in developing quantum Bayes’ rule Leifer and Spekkens 2013 ; Tsang 2022 ; Parzygnat and Buscemi 2023 ; Parzygnat and Fullwood 2023 ; Cenxin et al. 2023 ; Scandi et al. 2025 ; Surace and Scandi 2023 ; Bai et al. 2025 ; Liu et al. 2025a ; Liu et al. 2025b ; Liu et al. 2025c . Apart from its fundamental importance, the implementation of Petz recovery maps is substantial to make it practically useful. There have been a few developments toward this direction. The first attempt to implement Petz recovery maps approximately was given in Ref. Gilyén et al. 2022 . Later, circuit-design approaches have been proposed based on the resource availability in experimental setups (e.g. the circuit depth or the number of ancilla qubits) in the context of quantum error correction focusing on a code space Biswas et al. 2024 . Subsequently, an experimental realization of Petz recovery maps has been recently developed for a qubit decoherence channel by designing exact quantum circuits in trapped ion platforms Png and Scarani 2025 , and in NMR platforms Singh et al. 2025 . Here, we approach this problem in a different perspective— we ask when a Petz recovery map can be implemented without requiring additional devices than the existing ones on the table. Specifically, we ask when a Petz recovery map can be realized with the same devices that were available for the forward dynamics, thus enabling ‘cost-free’ but still nearly optimal reversal. This approach was first introduced in Ref. Aw et al. 2024 , and the forward dynamics is said to be tabletop time-reversible (TTR) when such a realization is possible. In this work, we study conditions of TTR in detail. We present the exact conditions of TTR in a directly usable form and we demonstrate it with illustrative examples. These conditions provide an insightful connection with the sufficient condition of TTR, product preservation, previously found in Ref. Aw et al. 2024 —it provides a straightforward explanation of why the product preservation is a strictly sufficient, but not necessary condition. Yet, the exact TTR conditions require a control of an ancilla system in a timely manner in general. We thus turn to approximate TTR conditions by which we circumvent such a time-sensitive control and present the first two non-trivial approximate TTR conditions; first-order approximate TTR and second-order approximate TTR. While such approximate TTR conditions contain complex terms in general, we present simpler forms of approximate TTR conditions for some special choices of a reference state. This includes the first-order approximate implementation of the class of Petz recovery maps relevant to quantum error correction. In addition, we present Lindbladian conditions of approximate TTR, with which one can implement Petz recovery maps sequentially. TTR-based approaches enable the implementation of Petz recovery maps, which does not require excessive resources as the unitary operations for the reversal are considered accessible and require only at most the same number of ancilla qubits as that of ancilla for the forward dynamics. II Background and Definitions (a) Dilation of Quantum Dynamics (b) Tabletop Time-reversible Dynamics Figure

Read the complete text