Pith. sign in

REVIEW 50 references

A post-selected fluctuation theorem yields a nonnegative entropy production for non-Hermitian quantum dynamics.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

T0 review · grok-4.5

2026-07-31 04:53 UTC pith:4BLJDO4R

load-bearing objection Solid first-principles entropy production for post-selected NH dynamics; the FT→Σ=D₂−D chain and EP coherent split are clean and correctly scoped.

arxiv 2607.24961 v1 pith:4BLJDO4R submitted 2026-07-27 quant-ph cond-mat.mes-hallcond-mat.stat-mech

Non-Hermitian entropy production from fluctuation theorems

classification quant-ph cond-mat.mes-hallcond-mat.stat-mech
keywords non-Hermitian dynamicsfluctuation theoremsentropy productionpost-selectionPetz-Rényi divergenceexceptional pointsquantum trajectories
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

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

Non-Hermitian evolution is an effective description of quantum systems conditioned on never jumping into the environment. This paper builds a thermodynamic account of that conditioned dynamics from first principles by writing a fluctuation theorem only for the no-jump trajectories. The resulting average entropy production stays nonnegative for the entire evolution and therefore supplies a second law for the post-selected process. The production splits into the ordinary average change of system entropy plus a strictly positive non-Hermitian correction that equals a Petz–Rényi divergence; the same objects also give tight upper and lower bounds. Because the correction is sensitive to non-commutativity of the Hermitian and anti-Hermitian parts of the effective Hamiltonian, its coherent piece carries a thermodynamic signature of exceptional points. A two-level model is used to show the signature explicitly.

Core claim

Starting from the ordinary system-bath fluctuation theorem and restricting to the zero-temperature no-jump Kraus operator, the authors obtain a normalized relation whose exponent defines a stochastic entropy production. Its average Σ = ⟨ΔS⟩ + Ξ is nonnegative throughout the non-Hermitian evolution, with Ξ = log(Tr(Q²ρ)/Tr(Qρ)²) equal to the order-2 Petz–Rényi divergence D₂(σ∥ρ) and ⟨ΔS⟩ = −D(σ∥ρ). Hence Σ itself is the difference of two Rényi divergences and saturates only in the trivial Hermitian limit.

What carries the argument

The post-selected fluctuation theorem P(ΔS)/P̃(−ΔS) = e^{ΔS+Ξ}, where the non-Hermitian correction Ξ is built from the operator Q = U_eff† U_eff and equals both the inverse squared signal-to-noise ratio of Q and the Petz–Rényi divergence D₂(σ∥ρ).

Load-bearing premise

The whole construction applies only to the zero-temperature no-jump trajectory and treats the backward state as the normalized final state of that trajectory, so the resulting entropy production has no direct meaning for the underlying unconditioned open system.

What would settle it

Prepare a qubit in a mixed state aligned with the exceptional-point eigenvector of the effective Hamiltonian, post-select the no-click trajectory, and measure whether the coherent part of Ξ develops a sharp negative dip exactly at the exceptional-point parameters while the total Σ remains nonnegative.

Watch this falsifier — get emailed when new claim-graph text bears on it.

If this is right

  • Post-selected non-Hermitian dynamics obeys a second law Σ ≥ 0 that is invariant under overall imaginary shifts of the effective Hamiltonian.
  • The non-Hermitian correction admits state-independent and state-dependent bounds expressed solely in terms of the extreme eigenvalues of Q and the max-relative entropy.
  • Coherent and incoherent pieces of Ξ can be separated; the coherent piece is nonzero only for non-normal Hamiltonians and can diagnose exceptional points.
  • The same identities hold for time-dependent effective Hamiltonians, opening a route to exceptional-point encircling and non-Markovian post-selection.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • Because Σ ignores heat exchange with the bath, any experimental test must compare only post-selected trajectory statistics, not the full open-system heat balance.
  • The rapid rotation of the Q-eigenaxis near an exceptional point suggests that Ξ_c could serve as a thermodynamic witness of exceptional-point sensitivity in sensing protocols.
  • Extending the post-selection to finite-temperature or multi-jump conditioned trajectories would test how much of the Rényi structure survives beyond the pure no-jump case.

Editorial analysis

A structured set of objections, weighed in public.

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

Circularity Check

0 steps flagged

No significant circularity: Σ ≥ 0 follows from post-selected FT normalization plus Petz monotonicity, not from self-definition or fitted inputs.

full rationale

The central claim is obtained by specializing the standard Kraus microreversibility relation to the zero-temperature no-jump channel, renormalizing the unnormalized no-jump weights into probabilities, and applying Jensen (or equivalently Petz-Rényi monotonicity in α). Appendix B computes N and Ñ explicitly from the definitions, yielding Ξ = log(Tr(Q²ρ)/Tr(Qρ)²) and ⟨ΔS⟩ = −D(σ∥ρ) with σ constructed from Q and ρ; positivity of Σ = D₂(σ∥ρ) − D(σ∥ρ) is then a standard inequality, not an identity forced by renaming the target. The choice ρ̃ = ρ(t) is conventional and stated openly; the paper explicitly scopes Σ to post-selected dynamics and disclaims a direct link to pre-selected open-system entropy production. Self-citations are background (prior NH entropy notions, trajectory FTs) and are not load-bearing for the positivity proof. The EP/coherent decomposition is an illustration, not part of the derivation of Σ ≥ 0. No fitted parameters, uniqueness theorems imported from the authors, or ansatz smuggling appear in the chain.

Axiom & Free-Parameter Ledger

0 free parameters · 5 axioms · 2 invented entities

The central positivity result rests on standard open-system microreversibility, the Kraus representation of a thermal channel, post-selection onto the no-jump trajectory, and the conventional choice of backward state equal to the normalized final state. No fitted parameters enter Σ. The only paper-specific modeling choices are the zero-temperature no-jump restriction and the coherent/incoherent split defined by setting H = 0.

axioms (5)
  • domain assumption Microreversibility of system-bath unitary evolution implies the Kraus relation M̃_μν = M†_νμ exp(β ω_μν / 2).
    Invoked in Preliminaries and Eq. (2); standard in quantum FT literature but not proved in the paper.
  • domain assumption Non-Hermitian evolution is exactly the post-selected no-jump trajectory (μ = ν = 0) of a zero-temperature environment.
    Stated after Eq. (6); required to drop heat terms and obtain Eq. (8).
  • ad hoc to paper Backward initial state is chosen as the normalized final state ρ̃ = ρ(t).
    Explicitly fixed after Eq. (10); other choices would redefine the entropy production variant.
  • standard math Monotonicity of Petz–Rényi divergences in the order α (D₂(σ∥ρ) ≥ D(σ∥ρ)).
    Used for Eq. (19); cited to Hiai (2024).
  • ad hoc to paper Incoherent contribution is defined by setting the Hermitian part H = 0 in H_eff = H − iF.
    Definition before Eq. (23); natural when [H,F]=0 but still a modeling choice for the split.
invented entities (2)
  • Non-Hermitian entropy production Σ := ⟨ΔS⟩ + Ξ independent evidence
    purpose: Provide a nonnegative, FT-derived irreversibility measure for post-selected non-Hermitian dynamics.
    Defined in Eqs. (9)–(10) from the normalized post-selected FT; not previously standard.
  • Coherent non-Hermitian correction Ξ_c = Ξ − Ξ_i independent evidence
    purpose: Isolate the contribution due to non-normality and diagnose exceptional points.
    Introduced in Eqs. (23)–(25); falsifiable via parameter scans near EPs as in Fig. 2.

pith-pipeline@v1.2.0-grok45-kimik3 · 17712 in / 3041 out tokens · 61344 ms · 2026-07-31T04:53:01.746092+00:00 · methodology

0 comments
read the original abstract

We develop a first-principles thermodynamic framework for non-Hermitian dynamics based on a post-selected version of the fluctuation theorem. This allows us to identify a quantity that remains positive throughout the non-Hermitian evolution and can be interpreted as the entropy production of the post-selected dynamics. We relate this quantity to previously proposed notions of non-Hermitian entropy and derive an associated second law. Furthermore, we establish a connection with information-theoretic quantities, in particular the Petz-R\'enyi divergences, and leverage this connection to derive upper and lower bounds. Finally, we decompose the entropy production into incoherent and coherent contributions, identifying distinctive features of the coherent term in the vicinity of exceptional points. We illustrate our results using a paradigmatic model of non-Hermitian evolution based on a two-level system.

Figures

Figures reproduced from arXiv: 2607.24961 by Donato Farina, Frank Ernesto Quintela Rodr\'iguez, Vasco Cavina.

Figure 1
Figure 1. Figure 1: FIG. 1: Σ( [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: FIG. 2: Coherent contribution Ξ [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

50 extracted references · 9 linked inside Pith

  1. [1]

    Ashida, Z

    Y. Ashida, Z. Gong, and M. Ueda, Advances in Physics 69, 249 (2020)

  2. [2]

    Xue, Phys

    P. Xue, Phys. Rev. Lett.136, 170001 (2026)

  3. [3]

    W. D. Heiss, Journal of Physics A: Mathematical and Theoretical45, 444016 (2012)

  4. [4]

    Minganti, A

    F. Minganti, A. Miranowicz, R. W. Chhajlany, and F. Nori, Phys. Rev. A100, 062131 (2019)

  5. [5]

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

  6. [6]

    Naghiloo, M

    M. Naghiloo, M. Abbasi, Y. N. Joglekar, and K. W. Murch, Nature Physics15, 1232 (2019)

  7. [7]

    Di Colandrea, F

    F. Di Colandrea, F. Pavan, S. Bansal, P. Savarese, G. Di Bello, G. De Filippis, C. A. Perroni, D. Farina, and F. Cardano, arXiv preprint arXiv:2512.09870 (2025)

  8. [8]

    Kawabata, K

    K. Kawabata, K. Shiozaki, M. Ueda, and M. Sato, Phys. Rev. X9, 041015 (2019)

  9. [9]

    Z. Gong, Y. Ashida, K. Kawabata, K. Takasan, S. Hi- gashikawa, and M. Ueda, Phys. Rev. X8, 031079 (2018)

  10. [10]

    Schomerus, Phys

    H. Schomerus, Phys. Rev. Res.2, 013058 (2020)

  11. [11]

    E. J. Bergholtz, J. C. Budich, and F. K. Kunst, Rev. Mod. Phys.93, 015005 (2021)

  12. [12]

    Yao and Z

    S. Yao and Z. Wang, Phys. Rev. Lett.121, 086803 (2018)

  13. [13]

    Wiersig, Phys

    J. Wiersig, Phys. Rev. Lett.112, 203901 (2014)

  14. [14]

    Hodaei, A

    H. Hodaei, A. U. Hassan, S. Wittek, H. Garcia-Gracia, R. El-Ganainy, D. N. Christodoulides, and M. Kha- javikhan, Nature548, 187 (2017)

  15. [15]

    Chen, S ¸

    W. Chen, S ¸. Kaya ¨Ozdemir, G. Zhao, J. Wiersig, and L. Yang, Nature548, 192 (2017)

  16. [16]

    N. Zeng, T. Liu, K. Xia, Y.-R. Zhang, and F. Nori, Phys. Rev. Res.7, 043219 (2025)

  17. [17]

    Roccati and F

    F. Roccati and F. Balducci, arXiv preprint arXiv:2603.13526 (2026)

  18. [18]

    Sergi and K

    A. Sergi and K. G. Zloshchastiev, J. Stat. Mech.: Theory 6 Exp.2016(3), 033102

  19. [19]

    Di Bello, F

    G. Di Bello, F. Pavan, V. Cataudella, and D. Farina, arXiv preprint arXiv:2603.23011 (2026)

  20. [20]

    Tusun and F

    M. Tusun and F. Zhao, arXiv preprint arXiv:2606.22405 (2026)

  21. [21]

    Z. Yang, T. Liu, X. Tan, F. Zhu, and L. Luo, arXiv preprint arXiv:2605.10099 (2026)

  22. [22]

    C. Lan, L. Ma, and H. Yang, arXiv preprint arXiv:2606.13251 (2026)

  23. [23]

    Campisi, P

    M. Campisi, P. H¨ anggi, and P. Talkner, Rev. Mod. Phys. 83, 771 (2011)

  24. [24]

    Esposito, U

    M. Esposito, U. Harbola, and S. Mukamel, Rev. Mod. Phys.81, 1665 (2009)

  25. [25]

    Soret, V

    A. Soret, V. Cavina, and M. Esposito, Phys. Rev. A106, 062209 (2022)

  26. [26]

    Manzano and R

    G. Manzano and R. Zambrini, A VS Quantum Science4 (2022)

  27. [27]

    Albarelli and M

    F. Albarelli and M. G. Genoni, Physics Letters A494, 129260 (2024), arXiv:2312.13214 [quant-ph]

  28. [28]

    M. M. Wauters, P. Boschetto, E. Ballini, A. Biella, and P. Hauke, arXiv preprint arXiv:2606.27424 (2026)

  29. [29]

    Settimo and J

    F. Settimo and J. Piilo, arXiv preprint arXiv:2605.07797 (2026)

  30. [30]

    Settimo, K

    F. Settimo, K. Luoma, D. Chru´ sci´ nski, B. Vacchini, A. Smirne, and J. Piilo, Physical Review A113, 042444 (2026)

  31. [31]

    Ferri-Cort´ es, J

    M. Ferri-Cort´ es, J. A. Almanza-Marrero, R. L´ opez, R. Zambrini, and G. Manzano, Phys. Rev. Res.7, 013077 (2025)

  32. [32]

    Fiusa, P

    G. Fiusa, P. E. Harunari, A. S. Hegde, and G. T. Landi, arXiv preprint arXiv:2505.06208 (2025)

  33. [33]

    Breuer and F

    H.-P. Breuer and F. Petruccione,The Theory of Open Quantum Systems(Oxford University Press, 2002)

  34. [34]

    Manzano, J

    G. Manzano, J. M. Horowitz, and J. M. Parrondo, Phys- ical Review E92, 032129 (2015)

  35. [35]

    Manzano, J

    G. Manzano, J. M. Horowitz, and J. M. Parrondo, Phys- ical Review X8, 031037 (2018)

  36. [36]

    Notice that, differently from [35], we absorbed a system time-reversal operator Θ in the definition of the time- reversed process

  37. [37]

    Kurchan, A quantum fluctuation theorem (2000), arXiv:cond-mat/0007360 [cond-mat.stat-mech]

    J. Kurchan, A quantum fluctuation theorem (2000), arXiv:cond-mat/0007360 [cond-mat.stat-mech]

  38. [38]

    Talkner and P

    P. Talkner and P. H¨ anggi, Physical Review E93, 022131 (2016)

  39. [39]

    Andrieux and P

    D. Andrieux and P. Gaspard, Phys. Rev. Lett.100, 230404 (2008)

  40. [40]

    Esposito and C

    M. Esposito and C. Van den Broeck, Physical review let- ters104, 090601 (2010)

  41. [41]

    Thus ker(ρ)⊆ker(σ), which is equivalent to supp(σ)⊆supp(ρ)

    Indeed, for anyv∈ker(ρ) one hasρ 1/2v= 0, and there- foreσv∼ρ 1/2Qρ1/2v= 0. Thus ker(ρ)⊆ker(σ), which is equivalent to supp(σ)⊆supp(ρ)

  42. [42]

    Datta, IEEE Transactions on Information Theory55, 2816 (2009)

    N. Datta, IEEE Transactions on Information Theory55, 2816 (2009)

  43. [43]

    Gour,Resources of the Quantum World(Cambridge University Press, Cambridge, 2025)

    G. Gour,Resources of the Quantum World(Cambridge University Press, Cambridge, 2025)

  44. [44]

    Hiai, Acta Scientiarum Mathematicarum90, 527 (2024)

    F. Hiai, Acta Scientiarum Mathematicarum90, 527 (2024)

  45. [45]

    Baumgratz, M

    T. Baumgratz, M. Cramer, and M. B. Plenio, Phys. Rev. Lett.113, 140401 (2014)

  46. [46]

    Francica, F

    G. Francica, F. C. Binder, G. Guarnieri, M. T. Mitchison, J. Goold, and F. Plastina, Phys. Rev. Lett.125, 180603 (2020)

  47. [47]

    Figueira de Morisson Faria and A

    C. Figueira de Morisson Faria and A. Fring, J. Phys. A 39, 9269 (2006)

  48. [48]

    Doppler, A

    J. Doppler, A. A. Mailybaev, J. B¨ ohm, U. Kuhl, A. Girschik, F. Libisch, T. J. Milburn, P. Rabl, N. Moi- seyev, and S. Rotter, Nature537, 76 (2016)

  49. [49]

    D’Abbruzzo, V

    A. D’Abbruzzo, V. Giovannetti, and V. Cavina, Phys. Rev. Lett.135, 240401 (2025)

  50. [50]

    Pellitteri, V

    G. Pellitteri, V. Giovannetti, and V. Cavina, arXiv preprint arXiv:2602.21190 (2026). 7 Appendix A: Fluctuation theorem and entropy production Eq. (5) can be obtained directly by definingP ν→µ(∆S) and applying the symmetry (2). We have Pν→µ(∆S) := X nm Pµ→ν(n, m)δ(∆S−∆Snm) = X nm | ⟨˜n|Mµν |m⟩ |2pmδ(∆S−∆S nm).(A1) We can now use the symmetry (2) and the d...