Non-Markovian Exceptional Points by Interpolating Quantum Channels

Bei Zeng; Jensen Li; Wai Chun Wong

REVIEW 4 major objections 5 minor 1 cited by

Non-Markovian Exceptional Points by Interpolating Quantum Channels

T0 review · 4 major / 5 minor · reviewed 2026-08-06 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read Mixing two single-qubit quantum channels from different phases forces an exceptional point where eigenvalues and eigenvectors coalesce.

desk verdict The specific two-channel EP example is correct and the experiment supports it, but the abstract's general claim overreaches: endpoint interpolations can produce Diabolic Points, not EPs, and the non-Markovian label is asserted rather than proved. read the letter →

arxiv 2507.16049 v1 pith:WI3LXTTX submitted 2025-07-21 quant-ph physics.app-ph

Wai Chun Wong , Bei Zeng , Jensen Li This is my paper

classification quant-phphysics.app-ph

keywords quantumchannelsexceptionalpointsnon-Markoviandynamicssingle-qubitphasetransitionprocesstomographyNMRcomputerhigher-order

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper claims that every single-qubit quantum channel falls into one of two phases: its eigenvalues are either all real or include one complex-conjugate pair. When two channels from different phases are linearly interpolated, the transition between phases is marked by an exceptional point, a parameter value where two eigenvalues and their eigenvectors coalesce. The authors exhibit an explicit channel pair whose interpolated eigenvalues are ±√(p/2 − 1/4), coalescing at p = 1/2, and they confirm this on an NMR quantum computer with process fidelities above 93%. Because the interpolated channels are completely positive and trace-preserving at every step yet generally cannot be divided into a sequence of valid channels, these channel-level EPs lie outside the Markovian Lindblad description, which is why the paper labels them non-Markovian. Interpolating three channels yields a third-order exceptional point.

What carries the argument

The machinery is the Bloch-vector representation of a single-qubit channel, in which the superoperator becomes a real 4×4 matrix whose first row is (1,0,0,0) by trace preservation, and whose 3×3 distortion matrix E is real. Realness forces eigenvalues to appear in conjugate pairs, giving the two phases, and the complex-conjugation operator K replaces PT symmetry as the organizing symmetry. The interpolation E(p) = (1 − p)E1 + pE2 connects distortion matrices from the two phases linearly, and the coalescence of eigenvalues and eigenvectors at the phase boundary is the exceptional point. For the experiment, E(p) is decomposed into a convex combination of two simpler channels Q1(p), Q2(p), each implementable with one ancilla qubit, and the full superoperator is reconstructed by maximum-likelihood quantum process tomography.

What would settle it

Compute the Jordan normal form of E(p) exactly at the predicted boundary p = 1/2 for the E1/E2 pair; if the matrix is diagonalizable there, the claimed exceptional point does not exist. For the general recipe, find two single-qubit channels in different phases whose interpolation at the phase boundary has distinct eigenvectors for the merged eigenvalue; that single counterexample would disprove the claim that all such transitions are EPs.

Watch

Extended reading notes

Core claim

The central discovery is a two-phase classification of single-qubit quantum channels and a general way to force an exceptional point. Writing a channel as a real 4×4 matrix acting on Bloch coordinates {1, rx, ry, rz}, trace preservation fixes the first row, leaving a real 3×3 distortion matrix E whose eigenvalues are either all real or one real plus one complex-conjugate pair. The paper names these the K-exact and K-broken phases, where K is complex conjugation, and notes a structural analogy to, but not a reliance on, PT symmetry. Linear interpolation E(p) = (1 − p)E1 + pE2 between channels in different phases must cross a boundary where eigenvalues merge; for the explicit pair the eigenvalues are λ± = ±√(p/2 − 1/4) with eigenvectors v± = {0, p + 2λ±, 1 − p}, both coalescing at p = 1/2. The authors take this as evidence that the boundary is a channel exceptional point, and they verify empirically via quantum process tomography that the predicted eigenvalue flow is reproduced with fidelities above 93% across the whole interpolation range.

Load-bearing premise

The general recipe only yields an exceptional point if the eigenvalue merger at the phase boundary is a defective matrix where eigenvectors also coalesce; the paper verifies this for its two example pairs but gives no proof that it always happens rather than a diabolic point with distinct eigenvectors.

Editorial extensions

If this is right

Every single-qubit quantum channel sits in one of two eigenvalue phases, so interpolating between opposite phases yields a channel exceptional point without any gain-loss or PT-symmetry engineering.
Because the interpolated channels are CPTP yet generally non-divisible, the resulting EPs are genuinely non-Markovian, extending exceptional-point physics beyond Lindblad and Liouvillian descriptions.
The explicit E1/E2 pair realizes a second-order EP at p = 1/2, confirmed on an NMR quantum computer with fidelities above 93% for all interpolated channels.
Interpolating three channels produces EP lines in the parameter triangle and a third-order EP at (a1, a2, a3) ≈ (0.446, 0.322, 0.232), showing that the scheme scales to higher-order EPs.
Every Markovian Liouvillian EP induces a channel EP through exponentiation E = e^{LT}, so the channel framework subsumes Liouvillian EPs and extends to processes with no master-equation representation.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

If the two-phase structure persists in higher-dimensional channels, channels would be classified by the number of complex-conjugate eigenvalue pairs, with EP hypersurfaces separating phase regions; this could serve as a practical diagnostic of non-Markovianity.
The recipe only needs any two channels in different phases, so one could random-search the single-qubit channel space for EP-producing pairs, turning the paper's 'try another pair' suggestion into a systematic numerical method.
A sharper test of the non-Markovian claim would be to compute the divisibility of E(p) via its Choi matrix; the paper asserts non-divisibility but does not demonstrate it for its experimental channels.
Near the exceptional point, perturbations in p split the eigenvalues with a square-root law, suggesting the interpolated channel could act as a sensor that amplifies small channel differences, analogous to EP-enhanced sensing in wave systems.

Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

The concrete result is real: interpolating their two example channels gives an exact EP at p = 1/2, with eigenvalues ±sqrt(p/2 - 1/4) and coalescing eigenvectors, and the NMR experiment tracks both real and imaginary parts with good fidelity. That is a clean, reproducible demonstration of a channel-level exceptional point, which I have not seen done this way before. The EP3 from three-channel interpolation is also a nice addition, even without experimental confirmation. Credit where due: the channel-level definition, the K-exact/K-broken phase language, and the explicit Kraus representations are all sensible, and the paper does not hide the Diabolic Point possibility in the main text.

The soft spots are real but localized. The abstract says the transition between phases is marked by the presence of an EP, and that is false as stated. Take the completely depolarizing channel (E = 0) and a unitary z-rotation by pi/2: at p = 0 you get a triple eigenvalue zero with three independent eigenvectors, a diagonalizable degeneracy, i.e. a Diabolic Point. So not every interpolation between different phases yields an EP; the paper's own later caveat about DPs is accurate, but it undermines the 'systematic method' language. You can rescue a weaker claim: for interior phase transitions, the eigenvalue discriminant crossing a simple zero typically forces a Jordan block, hence an EP, but endpoint transitions fail. The paper gives no criterion to distinguish EP from DP.

Second, the non-Markovian claim in the title and introduction is not substantiated. The authors assert that interpolated channels are generally non-divisible, which would indicate non-Markovian dynamics, but I cannot find a proof or even a reference for that specific step. It is plausible, but the label 'non-Markovian exceptional points' is currently stronger than what is shown.

The experiment also only shows eigenvalue coalescence, not eigenvector coalescence. Since they reconstruct the full channel via QPT, the eigenvectors are available in principle; showing the phase rigidity or eigenvector overlap at p = 0.5 would have directly confirmed the EP. Minor in comparison to the overclaim, but worth flagging.

All that said, the core eigenvalue calculation is exact, the experiment is a genuine implementation, and the paper is honest enough to mention DPs. It deserves a serious referee: the channel-interpolation idea is novel and likely useful, but the general claims need to be tempered and the non-Markovian part either proved or cut. I would not cite it in my own work as it stands, but I would read a revised version.

Referee Report

4 major / 5 minor

Summary. The paper proposes that single-qubit quantum channels, represented as real 4x4 matrices on a vectorized density matrix, fall into two spectral phases: all-real eigenvalues or one real plus one complex-conjugate pair. It claims that linearly interpolating between channels from different phases produces an exceptional point (EP) at the phase transition, and it demonstrates this on the explicit pair E1 and E2 in Eq. (4), whose interpolated distortion matrix has eigenvalues lambda_plus/minus = ±sqrt(p/2 - 1/4) and coalescing eigenvectors at p = 1/2. The authors implement the two-channel interpolation on an NMR quantum computer, extract eigenvalues by quantum process tomography, and report process fidelities above 93%. They also sketch an extension to three channels, claiming a third-order EP at a specific point in the convex hull. The core two-channel eigenvalue calculation is exact and parameter-free, and the experimental data agree with the predicted spectral collapse, but the general claims about a systematic method, guaranteed EPs, and non-Markovian behavior are not fully supported.

Significance. If the general framework were established, it would give a simple, symmetry-free route to exceptional points in the most general description of single-qubit open-system evolution, going beyond Lindblad and Hamiltonian approaches. The paper's strengths are its explicit analytic example (Eq. (6)) with no fitted parameters, the concrete CPTP Kraus representations, and the NMR experiment that verifies the predicted spectral degeneracy with high fidelity. However, the advertised generality is currently stronger than what is proved: the manuscript itself concedes that Diabolic Points can occur, and the non-Markovian characterization rests on an unproved divisibility assertion. The work is therefore a promising contribution whose central example is sound, but whose broader claims need re-scoping or additional proof.

major comments (4)

[Abstract and paragraph after Eq. (6)] The abstract and the text after Eq. (6) claim that interpolating any two quantum channels from different phases inevitably produces an exceptional point at the phase transition. This is false for endpoint transitions: take E1 as the completely depolarizing channel (E1 = 0, eigenvalues {1,0,0,0}, all real) and E2 as the unitary z-rotation by pi/2 (E2 = R_z(pi/2), eigenvalues {1,i,-i}, complex pair). The interpolation gives E(p) = p R_z(pi/2) with eigenvalues {1,p,±ip}; at p = 0 the zero eigenvalue has geometric multiplicity 3, so the degeneracy is a diagonalizable Diabolic Point, not an EP. Since the manuscript acknowledges DPs but gives no criterion for when the transition is defective, the 'systematic method' is not guaranteed. Please revise the general claim, e.g., by restricting to interior transitions where the discriminant changes sign simply and a Jordan block is forced, and adjust the abstract and title accordingly.
[Introduction and discussion after Eq. (2)] The title and introduction advertise 'non-Markovian' exceptional points, but the non-Markovian characterization is asserted without proof. The text says the interpolated channels 'are generally non-divisible, indicating non-Markovian dynamics' and identifies Markovian evolution with E = exp(LT), yet no divisibility analysis of E(p) is provided. Please either prove that the interpolated family is not CP-divisible (for example, by exhibiting a non-CP intermediate map in a decomposition of the channel) or soften the non-Markovian claim so that it is not load-bearing for the paper's conclusions.
[Three-channel extension, text near Fig. 4] The third-order EP claim rests on 'phase rigidity' calculations that are not shown: no definition of phase rigidity for channel superoperators, no formula, and no numerical values are given. The statement that the phase transition lines are confirmed to be EPs 'as at least one phase rigidity vanishes' is therefore not verifiable from the manuscript. Please include the phase-rigidity calculation or an equivalent Jordan-form criterion, and give enough detail to reproduce the claimed EP3 location (a1:a2:a3) = (10:2√13:3√3).
[Experimental section, Figs. 2 and 3] The experiment demonstrates coalescence of eigenvalues (Fig. 3) but not eigenvector coalescence, which is the defining feature of an exceptional point. No phase rigidity, eigenvector overlap, or Jordan-block indicator is reported for the reconstructed channels. To claim experimental confirmation of an EP rather than a mere spectral degeneracy, please extract and report a measure of defective eigenvector structure from the process-tomography data, or explicitly state that the experiment verifies only the predicted spectral collapse while the eigenvector coalescence is taken from theory.

minor comments (5)

[After Eq. (2)] There is a typo: 'To revel two distinct phases' should read 'To reveal two distinct phases'.
[Three-channel extension, text near Eq. (8)] The phrase 'we chosen to have the distortion matrix E3 as a rotation matrix' should read 'we choose the distortion matrix E3 to be a rotation matrix'; also the symbol E3 is used for both the channel and its distortion matrix, which is confusing.
[Fig. 4 caption and surrounding text] The values are inconsistent: the text says 'When ax decrease from 0.361 to 0.321', while the caption lists fixed a2 = {0.362, 0.322, 0.282}; 'ax' should be 'a1', and the numerical values should be harmonized.
[Appendix references] The text refers to Appendix A (circuit-parameter algorithm) and Appendix B (fitting and fidelity definition), but these appendices are not present in the manuscript as provided; please include them or remove the references so the experimental procedure is self-contained.
[References] Reference [41] (Ruskai, Szarek, and Werner) appears in the bibliography but is not cited in the body of the paper; please cite it where it is intended to support the CPTP or eigenvalue-phase discussion, or remove it.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity; the exceptional point is obtained from a direct, parameter-free eigenvalue calculation.

full rationale

The central result is self-contained. The paper computes the distortion matrix of the interpolated channel by linearity (Eq. 6), then directly solves for eigenvalues λ± = ±√(p/2 − 1/4) and eigenvectors v± = {0, p + 2λ±, 1 − p}, showing both coalesce at p = 1/2. This is an explicit calculation, not a fit, not an imported theorem, and not a quantity defined in terms of the conclusion. The two-phase classification (all-real vs. one complex-conjugate pair) follows from the real 4×4 matrix representation of the channel in the Bloch basis (Eq. 2), which is an invertible linear change of basis preserving eigenvalues; this is elementary real-matrix spectral theory, not a circular input. The paper also acknowledges Diabolic Points where only eigenvalues merge, so it does not assert that every phase transition is automatically an EP; the proposed 'systematic method' is a search heuristic rather than a forced prediction. Experimental process tomography uses maximum-likelihood channel reconstruction and compares the resulting eigenvalues to the analytic prediction; the EP location is not extracted from the fitted data. Self-citations appearing in the references are background context and are not load-bearing for the derivation. The non-Markovian/non-divisibility identification is asserted rather than proven, but that is an unsupported assumption or limitation, not a circular reduction.

Assumptions & free parameters 0 free parameters · 5 assumptions · 0 invented entities

The central example is a parameter-free eigenvalue calculation with no fitted constants. The main unstated burden is the generic claim that phase-boundary crossings are exceptional points rather than Diabolic Points, plus the unproved non-Markovian characterization of the interpolated channels.

assumptions (5)

standard math Every real 3x3 matrix has eigenvalues either all real or one real plus a complex-conjugate pair.
Invoked implicitly in the phase classification after Eq. (2).
standard math Single-qubit CPTP channels can be represented by a real 4x4 matrix in the Pauli basis with first row (1,0,0,0).
Standard Pauli transfer matrix representation used throughout, e.g., Eqs. (1)-(2).
domain assumption Convex combinations of CPTP maps are CPTP.
Justifies that E(p) = (1-p)E1 + pE2 is a valid quantum channel.
ad hoc to paper The phase boundary between all-real and complex-conjugate eigenvalue sectors is an exceptional point, or at least a potential one.
General assertion after Eq. (2) without a proof; the paper itself notes Diabolic Points as an alternative, so this is a load-bearing unproved premise.
ad hoc to paper The interpolated channels E(p) are non-divisible and therefore non-Markovian.
Asserted in the introduction; no divisibility analysis is provided, and the title and framing depend on this claim.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Non-Markovian Exceptional Points by Interpolating Quantum Channels." pith.science (2026). https://pith.science/paper/WI3LXTTX

@misc{pith2026250716049,
  author       = {Pith},
  title        = {Pith review of: Non-Markovian Exceptional Points by Interpolating Quantum Channels},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/WI3LXTTX}},
  note         = {Machine review of arXiv:2507.16049}
}

read the original abstract

Exceptional points (EPs) are special points in non-Hermitian systems where both eigenvalues and eigenvectors coalesce. In open quantum systems, these points are typically analyzed using effective non-Hermitian Hamiltonians or Liouvillian superoperators. While quantum channels offer the most general framework for describing state evolution in such systems, the existence and properties of EPs within this setting remain largely unexplored. In this work, we present a general strategy for generating quantum EPs for a single-qubit setting. We show that quantum channels can be separated into two distinct phases, with the transition between them marked by the presence of an EP. Based on this, we propose a systematic method to realize EPs by interpolating between quantum channels representing different phases. Experimentally, we implement these interpolated channels on a nuclear magnetic resonance (NMR) quantum computer and confirm the emergence of second-order EPs with high fidelity. Extending the interpolation to three channels further reveals third-order EPs. Our results establish quantum channel interpolation as a versatile framework for generating EPs and provide a general description of EPs in open quantum systems.

Figures

Figures reproduced from arXiv: 2507.16049 by the authors.

**Figure 2.** FIG. 2. (a) Circuit for the implementation of the quantum channels [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗

**Figure 3.** FIG. 3. (a) The real and (b) imaginary parts of the eigenvalues of the superoperator [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗

Figures from the paper (1 more)

**Figure 4.** Figure 4: FIG. 4. (a) Phase diagram for interpolating 3 quantum channels [PITH_FULL_IMAGE:figures/full_fig_p009_4.png]

Discussion (0). Sign in to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

Non-Markovian exceptional points in waveguide quantum electrodynamics
quant-ph 2026-04 unverdicted novelty 7.0 of 10

Non-Markovian exceptional points arise in waveguide QED, manifesting as transitions to oscillatory spontaneous emission in giant atoms with multiple coupling points and in collective emission of separated emitters.

Reference graph

Works this paper leans on

41 extracted references · 37 canonical work pages · cited by 1 Pith paper

[1]

Z. G. Yuto Ashida and M. Ueda, Advances in Physics 69, 249 (2020)

work page 2020
[2]

El -Ganainy, K

R. El -Ganainy, K. G. Makris, M. Khajavikhan, Z. H. Musslimani, S. Rotter, and D. N. Christodoulides, Nature Physics 14, 11 (2018)

work page 2018
[3]

H. M. Hurst and B. Flebus, Journal of Applied Physics 132, 220902 (2022)

work page 2022
[4]

Moiseyev, Non-Hermitian Quantum Mechanics (Cambridge University Press, 2011)

N. Moiseyev, Non-Hermitian Quantum Mechanics (Cambridge University Press, 2011)

2011
[5]

El-Ganainy, M

R. El-Ganainy, M. Khajavikhan, D. N. Christodoulides, and S. K. Ozdemir, Communications Physics 2, 37 (2019)

work page 2019
[6]

C. M. Bender and S. Boettcher, Phys. Rev. Lett. 80, 5243 (1998)

1998
[7]

S¸. K. Ozdemir, S. Rotter, F. Nori, and L. Yang, Nature Materials 18, 783 (2019)

work page 2019
[8]

L. Feng, R. El-Ganainy, and L. Ge, Nature Photonics 11, 752 (2017)

work page 2017

Show all 41 references

[9]

Merkel, V

A. Merkel, V . Romero-Garcıa, J. -P. Groby, J. Li, and J. Christensen, Physical Review B 98, 201102 (2018)

2018
[10]

Z. Lin, H. Ramezani, T. Eichelkraut, T. Kottos, H. Cao, and D. N. Christodoulides, Physical Review Letters 106, 213901 (2011)

2011
[11]

Yin and X

X. Yin and X. Zhang, Nature materials 12, 175 (2013)

2013
[12]

Peng, S¸

B. Peng, S¸. K. Ozdemir, M. Liertzer, W. Chen, J. Kramer, H. Yılmaz, J. Wiersig, S. Rotter, and L. Yang, Proceedings of the National Academy of Sciences 113, 6845 (2016)

2016
[13]

K. Ji, Q. Zhong, L. Ge, G. Beaudoin, I. Sagnes, F. Raineri, R. El-Ganainy, and A. M. Yacomotti, Nature Communications 14, 8304 (2023)

2023
[14]

W. R. Sweeney, C. W. Hsu, S. Rotter, and A. D. Stone, Physical review letters 122, 093901 (2019)

2019
[15]

C. Wang, W. R. Sweeney, A. D. Stone, and L. Yang, Science 373, 1261 (2021)

2021
[16]

Horner, L

H. Horner, L. Wild, Y . Slobodkin, G. Weinberg, O. Katz, and S. Rotter, Physical Review Letters 133, 173801 (2024)

2024
[17]

Slobodkin, G

Y . Slobodkin, G. Weinberg, H. Horner, K. Pichler, S. Rotter, and O. Katz, Science 377, 995 (2022)

2022
[18]

Chen, S¸

W. Chen, S¸. Kaya Ozdemir, G. Zhao, J. Wiersig, and L. Yang, Nature 548, 192 (2017)

2017
[19]

Zhang, W

M. Zhang, W. Sweeney, C. W. Hsu, L. Yang, A. Stone, and L. Jiang, Physical Review Letters 123, 180501 (2019)

2019
[20]

Kononchuk, J

R. Kononchuk, J. Cai, F. Ellis, R. Thevamaran, and T. Kottos, Nature 607, 697 (2022)

2022
[21]

Dembowski, B

C. Dembowski, B. Dietz, H. -D. Graf, H. Harney, A. Heine, W. Heiss, and A. Richter, Physical Review E—Statistical, Nonlinear, and Soft Matter Physics 69, 056216 (2004)

2004
[22]

Doppler, A

J. Doppler, A. A. Mailybaev, J. Bohm, U. Kuhl, A. Girschik, F. Libisch, T. J. Milburn, P. Rabl, N. Moiseyev, and S. Rotter, Nature 537, 76 (2016)

2016
[23]

L. Ding, K. Shi, Q. Zhang, D. Shen, X. Zhang, and W. Zhang, Physical Review Letters 126, 083604 (2021). 11

2021
[24]

Z. Ren, D. Liu, E. Zhao, C. He, K. K. Pak, J. Li, and G.-B. Jo, Nature Physics 18, 385 (2022)

2022
[25]

E. Zhao, Z. Wang, C. He, T. F. J. Poon, K. K. Pak, Y .-J. Liu, P. Ren, X. -J. Liu, and G. -B. Jo, Nature 637, 565 (2025)

2025
[26]

Bu, J.-Q

J.-T. Bu, J.-Q. Zhang, G.-Y . Ding, J.-C. Li, J.-W. Zhang, B. Wang, W.-Q. Ding, W.-F. Yuan, L. Chen, i. m. c. K. Ozdemir, F. Zhou, H. Jing, and M. Feng, Phys. Rev. Lett. 130, 110402 (2023)

2023
[27]

Zhang, J.-Q

J.-W. Zhang, J.-Q. Zhang, G.-Y . Ding, J.-C. Li, J.-T. Bu, B. Wang, L.-L. Yan, S.-L. Su, L. Chen, F. Nori, et al., Nature Communications 13, 6225 (2022)

2022
[28]

Sun and W

K. Sun and W. Yi, Phys. Rev. A 108, 013302 (2023)

2023
[29]

Minganti, A

F. Minganti, A. Miranowicz, R. W. Chhajlany, and F. Nori, Phys. Rev. A 100, 062131 (2019)

2019
[30]

Zhou, X.-D

Y.-L. Zhou, X.-D. Yu, C.-W. Wu, X.-Q. Li, J. Zhang, W. Li, and P.-X. Chen, Phys. Rev. Res. 5, 043036 (2023)

2023
[31]

Khandelwal, N

S. Khandelwal, N. Brunner, and G. Haack, PRX Quantum 2, 040346 (2021)

2021
[32]

Sun and W

K. Sun and W. Yi, AAPPS Bulletin 34, 22 (2024)

2024
[33]

W. Chen, M. Abbasi, B. Ha, S. Erdamar, Y . N. Joglekar, and K. W. Murch, Phys. Rev. Lett. 128, 110402 (2022)

2022
[34]

S. Abo, P. Tulewicz, K. Bartkiewicz, S¸. K. Ozdemir,¨ and A. Miranowicz, New Journal of Physics 26, 123032 (2024)

2024
[35]

Liu, P.-C

J.-D. Liu, P.-C. Kuo, N. Lambert, A. Miranowicz, F. Nori, Y .-H. Chen, Nature Communications 16, 1289 (2025)

2025
[36]

M. A. Nielsen and I. L. Chuang, Quantum computation and quantum information (Cambridge university press, 2010)

2010
[37]

W. F. Stinespring, Proc. Amer. Math. Soc. 6, 211 (1955)

1955
[38]

D.-S. Wang, D. W. Berry, M. C. de Oliveira, and B. C. Sanders, Phys. Rev. Lett. 111, 130504 (2013)

2013
[39]

I. L. Chuang and M. A. Nielsen, Journal of Modern Optics 44, 2455 (1997)

1997
[40]

Gilchrist, N

A. Gilchrist, N. K. Langford, and M. A. Nielsen, Phys. Rev. A 71, 062310 (2005)

2005
[41]

M. B. Ruskai, S. Szarek, and E. Werner, Linear algebra and its applications 347, 159 (2002)

2002

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.

[1] [1]

Z. G. Yuto Ashida and M. Ueda, Advances in Physics 69, 249 (2020)

work page 2020

[2] [2]

El -Ganainy, K

R. El -Ganainy, K. G. Makris, M. Khajavikhan, Z. H. Musslimani, S. Rotter, and D. N. Christodoulides, Nature Physics 14, 11 (2018)

work page 2018

[3] [3]

H. M. Hurst and B. Flebus, Journal of Applied Physics 132, 220902 (2022)

work page 2022

[4] [4]

Moiseyev, Non-Hermitian Quantum Mechanics (Cambridge University Press, 2011)

N. Moiseyev, Non-Hermitian Quantum Mechanics (Cambridge University Press, 2011)

2011

[5] [5]

El-Ganainy, M

R. El-Ganainy, M. Khajavikhan, D. N. Christodoulides, and S. K. Ozdemir, Communications Physics 2, 37 (2019)

work page 2019

[6] [6]

C. M. Bender and S. Boettcher, Phys. Rev. Lett. 80, 5243 (1998)

1998

[7] [7]

S¸. K. Ozdemir, S. Rotter, F. Nori, and L. Yang, Nature Materials 18, 783 (2019)

work page 2019

[8] [8]

L. Feng, R. El-Ganainy, and L. Ge, Nature Photonics 11, 752 (2017)

work page 2017