Exact and variational identities for free energy differences in strongly coupled open systems

Christopher J. Stein; Mohammad Rahbar

REVIEW 5 major objections 5 minor 73 references

Exact and variational identities for free energy differences in strongly coupled open systems

T0 review · 5 major / 5 minor · reviewed 2026-08-03 · deepseek-v4-flash

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

Pith's one-line read Exact free-energy identities for open systems that need no reversibility, detailed balance, or fluctuation-dissipation structure, with the Jarzynski equality as a limiting case.

desk verdict Correct but known HMF/FEP identities are dressed up as a new fluctuation theorem; the abstract promises variational results that the text never delivers. read the letter →

arxiv 2511.10236 v3 pith:UJZZTZTZ submitted 2025-11-13 cond-mat.stat-mech physics.chem-phphysics.data-an

Mohammad Rahbar , Christopher J. Stein This is my paper · ORCID

classification cond-mat.stat-mechphysics.chem-phphysics.data-an

keywords HamiltonianofmeanforceJarzynskiequalityfluctuationtheoremsnonequilibriumfreeenergystrongcouplingchi-squareddivergencenon-Liouvilliandynamicsopenquantumsystems

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 derives exact identities for the free energy difference between two equilibrium endpoints of an open system, valid for arbitrary driven dynamics that connect those endpoints. The exponential of the free-energy difference is expressed as an exponential moment of the Hamiltonian-of-mean-force shift divided by a chi-squared overlap between the endpoint marginals, and a complementary identity holds for the negative exponent. These endpoint identities require no microscopic reversibility, detailed balance, fluctuation-dissipation structure, or local detailed balance. A trajectory-level version is proven under an asymptotic equilibration postulate, and in the frozen-coupling regime the identities reduce to bare-system expressions with a heat-work-feedback decomposition. The significance is that free energy differences can in principle be reconstructed from nonequilibrium trajectories in regimes where the standard Jarzynski equality fails, such as strongly coupled, active, or biological systems.

What carries the argument

The central objects are the Hamiltonian of mean force (HMF) H*_β, which defines the open system's canonical marginal and free energy F*_S, and the chi-squared divergence χ²(P^S_teq∥P^S_0) between the endpoint marginals. The asymptotic equilibration postulate — that the time-evolved marginal converges to the final canonical HMF marginal — is the mechanism that converts the algebraic endpoint identities into trajectory-level equalities. In the frozen-coupling regime, the HMF shift reduces to the bare-system Hamiltonian shift, yielding a heat-work-feedback decomposition (W_S, Q_S, and a reference functional II) that compensates dynamical asymmetry.

What would settle it

Run a driven open system with the same endpoints but a protocol that demonstrably relaxes to a non-canonical non-equilibrium steady state (e.g., an active-matter bath or an underdamped Langevin system with a non-conservative force). If the trajectory equality e^{+βΔF*_S}=⟨e^{+βΔH*_β}⟩^S still holds numerically, the asymptotic equilibration postulate is not necessary; if it fails, the postulate is confirmed as the limiting condition. A direct check is to measure the final marginal P^S_t at large t and compare it to the HMF canonical form via the chi-squared divergence — nonzero divergence signa

Watch

Extended reading notes

Core claim

The central claim is that for an open system S coupled to an environment E, with equilibrium endpoints at inverse temperature β described by Hamiltonian-of-mean-force (HMF) free energies F*_S, the identities e^{-βΔF*_S} = ⟨e^{-βΔH*_β}⟩^S / (1+χ²(P^S_teq∥P^S_0)) and e^{+βΔF*_S} = ⟨e^{+βΔH*_β}⟩^S hold, where the average is over the final HMF marginal and χ² is the chi-squared divergence between endpoint marginals. These endpoint equalities are proven without any dynamical generator assumptions; they follow from equilibrium partition-function ratios. Under the asymptotic equilibration postulate — that the driven final marginal equals the canonical HMF marginal — the equalities acquire exact tra

Load-bearing premise

The load-bearing premise is Eq. (19): that after the driving stops and the composite relaxes, the final marginal of the open system is exactly the canonical Hamiltonian-of-mean-force marginal at the final control parameters.

Editorial extensions

If this is right

If correct, the identities provide a route to reconstruct free energy differences from nonequilibrium trajectories in strongly coupled, active, or biological systems where detailed balance and local detailed balance are violated.
The standard Jarzynski equality and the Sagawa-Ueda generalized Jarzynski equality become special cases, so the framework unifies and enriches them by explicitly exposing endpoint overlap and system-environment interaction effects.
In the frozen-coupling regime, the trajectory equalities reduce to bare-system expressions, allowing practical estimation using only system coordinates and a heat-work-feedback correction.
The chi-squared overlap term acts as a diagnostic: large overlap burden signals poor endpoint overlap, which explains common failures of the Jarzynski equality in simulations and experiments.
The variational scalar-action law (maximum-entropy construction) is claimed to provide three outputs from the same sampled configurations — the HMF free energy difference, the endpoint-overlap burden, and a Hessian uncertainty estimate — independent of the microscopic system and environment at the level of the variational reconstruction.

Reading between the lines

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

The endpoint equalities (Theorem 1) are essentially equilibrium free-energy perturbation identities rewritten in HMF variables; their content is algebraic rather than dynamical. The real dynamical claim is the trajectory form (Theorem 2), which stands or falls on the asymptotic equilibration postulate.
A testable extension: for protocols that relax to a non-canonical non-equilibrium steady state, the trajectory identities should fail exactly when Eq. (19) is violated; quantifying the discrepancy would measure how far the final marginal is from the HMF canonical form.
The chi-squared overlap term can be viewed as a sampling-efficiency pre-factor: its divergence signals the regime where exponential averaging becomes dominated by rare trajectories, connecting to known difficulties with Jarzynski estimators in high-dimensional systems.
The heat-work-feedback decomposition suggests a practical protocol-design criterion: minimizing the reference functional II relative to the realized heat Q_S should reduce the variance of the exponential estimator, which could be tested in simulation.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

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 paper is a clean algebraic repackaging of free-energy perturbation in Hamiltonian-of-mean-force variables, with a trajectory extension that rests on an unproven and restrictive relaxation postulate, and a numerical validation that is essentially tautological. The abstract advertises a maximum-entropy variational construction, a Bessel-form scalar-action law, and Hessian uncertainty estimates; none of that appears in the full text.

What's good: the endpoint identities in Theorem 1 are derived correctly, the chi-squared overlap is a genuinely useful diagnostic for endpoint mismatch, and the heat-work-feedback decomposition in Corollary 2 is a neat algebraic separation, even if it's bookkeeping. The writing is clear, and the authors are honest that Theorem 1 coincides with their own prior endpoint equality [50]. The algebra checks out.

The problems are in the framing. Theorem 1 is exactly the standard Zwanzig FEP identity and its reverse/overlap form in HMF language—no dynamics, no fluctuation theorem. Theorem 2, the supposed 'beyond JE' content, requires Eq. (19), the asymptotic equilibration postulate: the final marginal must be exactly the canonical HMF marginal. That is not proven or characterized; it is deferred to a companion paper. For many advertised applications—active matter, biology—stationary marginals are non-Boltzmann, so the postulate fails. The validation builds the postulate in: the ramp is followed by underdamped Langevin relaxation whose stationary state is canonical, so comparing Eqs. (24)-(25) to the 'exact' partition ratio is checking the same identity against itself. The JE failure in their figures is a nice illustration of phase-space compression breaking Liouville, but it does not test the generality claim. The 'JE as limiting case' statement is true only in the trivial sense that Eq. (36) is the partition ratio; the appendix just recapitulates Jarzynski's own derivation, and the framework is not needed.

The abstract mismatch is the most serious issue: it promises a variational construction with Bessel-form and Hessian outputs that do not exist in the manuscript. That is not a minor editing slip.

Who is this for? A reader working on strong-coupling free energy estimation might take away the chi-squared overlap diagnostic. But as a Letter claiming to go beyond the Jarzynski equality, it does not deliver. My verdict: the math is right but the central claim is unsupported. A serious referee should see it because the issues are substantive and worth catching, but my own recommendation would be reject as a standalone Letter unless the variational results are actually included and the postulate is characterized.

Referee Report

5 major / 5 minor

Summary. The paper derives identities for the Hamiltonian of mean force (HMF) free energy difference ΔF*_S between two equilibrium endpoints of an open system, claiming they hold for arbitrary driven dynamics without microscopic reversibility, detailed balance, fluctuation-dissipation, or local detailed balance. Theorem 1 (Eqs. 15–16) expresses e^{±βΔF*_S} as exponential moments of the HMF shift averaged over the final system marginal, with a chi-squared endpoint-overlap term in one form. Theorem 2 (Eqs. 21–22) gives trajectory versions under an 'asymptotic equilibration' postulate (Eq. 19). Frozen-coupling corollaries (Eqs. 24–25) and a heat-work-feedback decomposition (Eqs. 26–34) are presented, and the paper claims the Jarzynski equality (JE) and its feedback extension (GJE) are recovered as limiting cases. A numerical validation on a non-Liouvillian ramp followed by underdamped Langevin relaxation is reported. The abstract also promises a maximum-entropy variational construction, which does not appear in the main text.

Significance. If the claims were fully established, the framework would be a significant step beyond traditional fluctuation theorems, because it would allow free-energy reconstruction for strongly coupled, non-reversible dynamics without detailed balance. The algebraic identities in Theorem 1 are correct, and the numerical demonstration that the naive JE estimator fails under a non-Liouvillian ramp while the endpoint identities reproduce the exact partition ratio is a useful consistency check. However, the central scientific claim is not supported. Theorem 1 is essentially a rewriting of known free-energy perturbation identities with no dynamical content; Theorem 2 depends on an unproven and restrictive equilibrium-relaxation postulate; and the numerical validation builds that postulate into the simulation protocol. The advertised applications to active matter and biological systems, whose stationary marginals are generally non-canonical, are therefore not justified by anything in the paper.

major comments (5)

[§3, Eq. (19) and Theorem 2] The trajectory counterparts (Eqs. 21–22) are valid only under the asymptotic equilibration postulate (19), which asserts that the final system marginal equals the HMF canonical marginal. This is not a consequence of 'arbitrary dynamics'; it is a strong equilibrium assumption. The paper does not prove sufficient conditions or characterize the protocols and dynamics for which it holds, and it defers to a companion paper [49]. For active-matter and biological systems, the stationary marginal is generally non-canonical (e.g., active Ornstein–Uhlenbeck particles), so the advertised generality is unsupported. Without Eq. (19), only Theorem 1 remains, which is an equilibrium identity with no dynamics.
[§2, Eqs. (15)–(16)] The endpoint equalities are algebraic identities: Eq. (16) is the standard free-energy perturbation formula ⟨e^{β(H*_f−H*_i)}⟩_{P_f} = Z_i/Z_f = e^{βΔF*}, and Eq. (15) is the corresponding overlap form. The trajectory map T_t and the dynamics do not enter these relations. Claiming that they constitute 'exact fluctuation relations' for nonequilibrium processes overstates their content: JE is a trajectory-level statement about work under a specified dynamical generator, while Theorem 1 requires sampling the exact equilibrium final marginal. The paper should explicitly state that Theorem 1 is an equilibrium identity, not a nonequilibrium fluctuation theorem.
[§4, JE as limiting case] Eqs. (35)–(40) do not derive JE as a limiting case of Theorem 1 or Theorem 2. They instead re-derive the conventional JE by assuming Hamiltonian dynamics, Liouville's theorem, and a work definition. Eq. (40) is simply the standard JE written in HMF notation. The statement that the framework 'includes the JE and its feedback extensions (GJE) as limiting cases' is unsupported; in particular, GJE is never derived or shown to reduce to a limit of the present identities, only mentioned qualitatively in the 'GJE connection' paragraph.
[§5, Validation and Fig. 1] The numerical validation is a consistency check, not a test of the trajectory theorem. The 'exact' reference is the partition ratio in Eq. (35), and the estimators in Eqs. (24)–(25) are evaluated after an underdamped Langevin relaxation that was deliberately chosen to satisfy Eq. (19) and converge to the same canonical ensemble used to compute the exact reference. Thus the agreement in Fig. 1 is built into the protocol. The failure of the JE estimator is expected because W_λ is not the proper work for the non-Liouvillian ramp; this does not demonstrate that the new identities are useful for arbitrary nonequilibrium processes. A meaningful test would need to verify Eq. (19) independently or use a protocol whose final stationary marginal is non-canonical.
[Abstract and main text] The abstract and title promise a 'maximum-entropy construction' yielding a 'Bessel-form scalar-action law' and 'Hessian uncertainty estimates,' but no such construction appears in the main text or the appendix. If this content is deferred to the companion paper [49], the present Letter is incomplete relative to its stated scope; if it is intended to be part of this work, it is missing. This is a substantive omission, as the 'variational' aspect of the claimed result is absent.

minor comments (5)

[Notation, Eqs. (15)–(20)] The notation ⟨•⟩^S for the final-marginal average and ⟨•⟩_{X0} for the initial full-system average is easy to confuse; consider using distinct bracket symbols or subscripts.
[Table I] The value C=1.0 is listed twice under 'System–environment coupling'; one entry appears to be a typo.
[Eq. (18)] The chi-squared divergence is written as −1 + ∫ (P_teq^S)^2 / P_0^S; please state explicitly that this is χ²(P_teq^S ∥ P_0^S) and define the orientation clearly.
[Corollary 2] The label 'feedback' for the term II (Eq. (31)) is misleading; this term is a projection of the driven velocity onto the initial force field. The connection to information-theoretic feedback in the Sagawa–Ueda sense is not established, and using the same word may cause confusion.
[Theorem 2 wording] The phrase 'exact trajectory forms' may suggest the identities hold for any trajectory ensemble; in fact they require the final pushed-forward marginal to equal the equilibrium HMF marginal. Please add an explicit caveat beneath Theorem 2.

Circularity Check

3 steps flagged · score 8.0 of 10

Central endpoint equalities are algebraic rewritings of the HMF partition-function ratio; trajectory forms and validation reduce to an imposed asymptotic-equilibration input.

self definitional [Theorem 1 (Eqs. 15-16); Section 'JE as limiting case' (Eqs. 35-36)]
"The central identity underlying all our derivations is e^{−β∆F*_S} = Z*(λ(t_eq), C(t_eq), β)/Z*(λ(0), C(0), β) ... By the definition of the HMF ... this equality ... contains no reference to the dynamics of the composite system. All relations derived in this work ... reduce to this equilibrium partition function ratio."

ΔF*_S is defined by the HMF partition functions (Eqs. 13-14). Eq. (16), e^{+βΔF*}=⟨e^{+βΔH*}⟩_f, is just the standard reverse free-energy perturbation identity Z_i/Z_f rewritten with the final HMF marginal, and Eq. (15) is the same ratio re-expressed with χ²(P_f∥P_i). No dynamical generator, reversibility condition, or independent physical input enters; the 'prediction' is the input partition-function ratio in new variables.
fitted input called prediction [Eq. (19) and 'Validation model and protocol' (Appendix)]
"We postulate asymptotic equilibration, lim_{t→t_eq} P^S_t(X_S,β) = P^S_t_eq(X_S,β) (19) ... These equations converge to the canonical distribution associated with the frozen Hamiltonian H^{S+E}(x,p_x,y,p_y;λ(t_eq),C), satisfying the asymptotic equilibration postulate Eq. (19)."

The trajectory forms (Thm. 2, Eqs. 21-22) become Eqs. (15)-(16) only when the pushforward of the initial ensemble is replaced by the final equilibrium HMF marginal. The validation chooses underdamped Langevin relaxation whose stationary state is exactly canonical, so Eq. (19) holds by construction. The agreement of the estimators (Eqs. 24-25) with the 'exact' partition ratio is therefore an internal algebraic consistency check, not an independent test of the claimed generality to non-canonical endpoints.

1 more flagged steps

self citation load bearing [Statement of Theorem 1 and Theorem 2; companion references [49], [50]]
"The result established here coincides with the endpoint equality derived in [50] ... The explicit pushforward expressions and a detailed discussion of this boundary condition are given in [49]."

The central theorems are not proved in this Letter; Theorem 1 is identified with the same-authors preprint [50], and the crucial asymptotic-equilibration boundary condition and the passage from Eqs. (15)-(16) to (21)-(22) are deferred to the same-authors companion [49]. Since [49] is a joint submission with this article and [50] is by the same authors, the load-bearing derivation chain reduces to self-citation rather than to independent, machine-checked, or externally reproduced evidence.

full rationale

The manuscript is internally consistent, and the endpoint equalities are mathematically correct identities, but they are definitional rather than derived from first-principles dynamics. Eq. (35) explicitly identifies the central content as the HMF partition-function ratio, and Theorem 1 re-expresses that ratio through endpoint marginals; Theorem 2 becomes exact only under the unproven asymptotic-equilibration postulate (Eq. 19), which the numerical example enforces by choosing a canonical-stationary relaxation. Thus the headline 'beyond Jarzynski' prediction—valid for arbitrary dynamics and applicable to active/biological systems—is not independently established: the algebraic part is FEP in HMF variables, and the trajectory generalization rests on an input condition that is assumed, not characterized. The χ² overlap and heat-work-feedback decomposition are useful rearrangements but do not supply independent physical content. This warrants a high circularity score because the claimed outputs reduce by construction to the inputs, though the identities themselves are not false.

Assumptions & free parameters 0 free parameters · 3 assumptions · 0 invented entities

The central derivation uses no fitted numerical parameters; the identities follow from the definitions of the HMF partition functions and the assumption that the system reaches the HMF-canonical final marginal. The validation model's parameters (Table I) are simulation inputs, not fitted. The key hidden input is the asymptotic-equilibration postulate.

assumptions (3)

domain assumption Asymptotic equilibration (Eq. 19): the final system marginal equals the HMF canonical marginal.
Postulated to pass from endpoint equalities to trajectory forms; not proven for the active/bio systems advertised in the conclusions.
domain assumption The composite system relaxes to the canonical state at the final fixed control parameters (Eqs. 3-6).
This is the standard equilibrium-bath assumption; it fails for non-equilibrium steady states, limiting the claimed scope.
standard math Definitions and standard math: canonical ensemble, HMF representation, chi-squared divergence, and interchange of integrals/limits in passing from Eqs. (15)-(16) to (21)-(22).
Exact identities follow from these definitions; no microscopic dynamics is used.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Exact and variational identities for free energy differences in strongly coupled open systems." pith.science (2026). https://pith.science/paper/UJZZTZTZ

@misc{pith2026251110236,
  author       = {Pith},
  title        = {Pith review of: Exact and variational identities for free energy differences in strongly coupled open systems},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/UJZZTZTZ}},
  note         = {Machine review of arXiv:2511.10236}
}

read the original abstract

We derive exact identities for open systems connecting two equilibrium endpoints without imposing microscopic reversibility, detailed balance (DB), fluctuation-dissipation structure, or local detailed balance (LDB) on the driven dynamics. The identities express the Hamiltonian of mean force (HMF) free energy differences through exponential moments and an explicit chi-squared overlap between the endpoint marginals. In the frozen-coupling regime, the HMF shift reduces to a bare-system increment and admits a trajectory-level heat-work-reference decomposition. The exact relations then reduce the problem to a scalar-action law. A maximum-entropy construction gives a Bessel-form scalar-action law, independent of the microscopic system, environment, and number of degrees of freedom at the level of the variational reconstruction. This law provides three outputs from the same sampled configurations: the HMF free energy difference, the endpoint-overlap burden, and a Hessian uncertainty estimate. Since many systems in biology, chemistry, physics and engineering violate the underlying assumptions of the standard Jarzynski identity, we validate the framework on a reduced-dimensional model with a non-Liouvillian, phase-space-compressing ramp followed by underdamped Langevin relaxation. The standard Jarzynski work estimator fails for this ramp because phase-space preservation is broken and no compensating Jacobian correction is included, whereas the present endpoint identities recover the exact HMF free energy difference, and the variational construction reproduces it within its local uncertainty.

Figures

Figures reproduced from arXiv: 2511.10236 by the authors.

**Figure 2.** ). Two-stage protocol. We work in the frozencoupling regime C(t) ≡ C(0) and vary only λ(t) (see [PITH_FULL_IMAGE:figures/full_fig_p008_2.png] view at source ↗

**Figure 3.** FIG. 3. Same plotting conventions as the Fig. 1. During the ramp we set [PITH_FULL_IMAGE:figures/full_fig_p009_3.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

73 extracted references · 3 linked inside Pith

[50]

Argun, A.-R

A. Argun, A.-R. Moradi, E. Pin¸ ce, G. B. Bagci, A. Im- parato, and G. Volpe, Non-Boltzmann stationary dis- tributions and nonequilibrium relations in active baths, Phys. Rev. E.94, 062150 (2016)

2016
[49]

Fodor, C

´E. Fodor, C. Nardini, M. E. Cates, J. Tailleur, P. Visco, and F. Van Wijland, How far from equilibrium is active matter?, Phys. Rev. Lett.117, 038103 (2016)

2016
[1]

Jarzynski, Nonequilibrium equality for free energy dif- ferences, Phys

C. Jarzynski, Nonequilibrium equality for free energy dif- ferences, Phys. Rev. Lett.78, 2690 (1997)

1997
[2]

Esposito, U

M. Esposito, U. Harbola, and S. Mukamel, Nonequilib- rium fluctuations, fluctuation theorems, and counting statistics in quantum systems, Rev. Mod. Phys.81, 1665 (2009)

2009
[3]

G. E. Crooks, Nonequilibrium measurements of free en- ergy differences for microscopically reversible Markovian systems, J. Stat. Phys.90, 1481 (1998)

1998
[4]

G. E. Crooks, Entropy production fluctuation theorem and the nonequilibrium work relation for free energy dif- ferences, Phys. Rev. E60, 2721 (1999)

1999
[5]

Liphardt, S

J. Liphardt, S. Dumont, S. B. Smith, I. Tinoco Jr, and C. Bustamante, Equilibrium information from nonequi- librium measurements in an experimental test of Jarzyn- ski’s equality, Science296, 1832 (2002)

2002
[6]

Ohzeki, Quantum annealing with the Jarzynski equal- ity, Phys

M. Ohzeki, Quantum annealing with the Jarzynski equal- ity, Phys. Rev. Lett.105, 050401 (2010)

2010

Show all 73 references

[7]

Xiong, L

T. Xiong, L. Yan, F. Zhou, K. Rehan, D. Liang, L. Chen, W. Yang, Z. Ma, M. Feng, and V. Vedral, Experimental verification of a Jarzynski-related information-theoretic equality by a single trapped ion, Phys. Rev. Lett.120, 6 010601 (2018)

2018
[8]

W. Liu, Z. Niu, W. Cheng, X. Li, C.-K. Duan, Z. Yin, X. Rong, and J. Du, Experimental test of the Jarzyn- ski equality in a single spin-1 system using high-fidelity single-shot readouts, Phys. Rev. Lett.131, 220401 (2023)

2023
[9]

N. C. Harris, Y. Song, and C.-H. Kiang, Experimental free energy surface reconstruction from single-molecule force spectroscopy using Jarzynski’s equality, Phys. Rev. Lett.99, 068101 (2007)

2007
[10]

Douarche, S

F. Douarche, S. Ciliberto, A. Petrosyan, and I. Rabbiosi, An experimental test of the Jarzynski equality in a me- chanical experiment, Europhys. Lett.70, 593 (2005)

2005
[11]

Saira, Y

O.-P. Saira, Y. Yoon, T. Tanttu, M. M¨ ott¨ onen, D. Averin, and J. P. Pekola, Test of the Jarzynski and Crooks fluctuation relations in an electronic system, Phys. Rev. Lett.109, 180601 (2012)

2012
[12]

Wimsatt, O.-P

G. Wimsatt, O.-P. Saira, A. B. Boyd, M. H. Math- eny, S. Han, M. L. Roukes, and J. P. Crutchfield, Har- nessing fluctuations in thermodynamic computing via time-reversal symmetries, Phys. Rev. Researh3, 033115 (2021)

2021
[13]

J. Gore, F. Ritort, and C. Bustamante, Bias and er- ror in estimates of equilibrium free-energy differences from nonequilibrium measurements, Proc. Natl. Acad. Sci. U.S.A.100, 12564 (2003)

2003
[14]

Toyabe, T

S. Toyabe, T. Sagawa, M. Ueda, E. Muneyuki, and M. Sano, Experimental demonstration of information- to-energy conversion and validation of the generalized Jarzynski equality, Nat. Phys.6, 988 (2010)

2010
[15]

J. Deng, J. D. Jaramillo, P. H¨ anggi, and J. Gong, De- formed Jarzynski equality, Entropy19, 419 (2017)

2017
[16]

Hern´ andez-G´ omez, S

S. Hern´ andez-G´ omez, S. Gherardini, A. Belenchia, A. Trombettoni, M. Paternostro, and N. Fabbri, Experi- mental signature of initial quantum coherence on entropy production, npj Quantum Information9, 86 (2023)

2023
[17]

D. Hahn, M. Dupont, M. Schmitt, D. J. Luitz, and M. Bukov, Quantum many-body Jarzynski equality and dissipative noise on a digital quantum computer, Phys. Rev. X.13, 041023 (2023)

2023
[18]

Press´ e and R

S. Press´ e and R. Silbey, Ordering of limits in the Jarzyn- ski equality, J. Chem. Phys.124(2006)

2006
[19]

Chen, On the Crooks fluctuation theorem and the Jarzynski equality, J

L. Chen, On the Crooks fluctuation theorem and the Jarzynski equality, J. Chem. Phys.129(2008)

2008
[20]

G. E. Crooks, Comment regarding” on the Crooks fluctu- ation theorem and the Jarzynski equality”[j. chem. phys. 129, 091101 (2008)] and” nonequilibrium fluctuation- dissipation theorem of Brownian dynamics”[j. chem. phys. 129, 144113 (2008)], J. Chem. Phys. , 107101 (2009)

2008
[21]

Sagawa and M

T. Sagawa and M. Ueda, Generalized Jarzynski equality under nonequilibrium feedback control, Phys. Rev. Lett. 104, 090602 (2010)

2010
[22]

Cohen and D

E. Cohen and D. Mauzerall, A note on the Jarzynski equality, J. Stat. Mech.: Theory Exp.2004(07), P07006

2004
[23]

Cohen and D

E. Cohen and D. Mauzerall, The Jarzynski equality and the Boltzmann factor, Mol. Phys.103, 2923 (2005)

2005
[24]

Argun, A.-R

A. Argun, A.-R. Moradi, E. Pince, G. B. Bagci, and G. Volpe, Experimental evidence of the failure of Jarzynski equality in active baths, arXiv preprint arXiv:1601.01123 (2016)

2016 arXiv
[25]

J. M. Vilar and J. M. Rubi, Failure of the work- Hamiltonian connection for free-energy calculations, Phys. Rev. Lett.100, 020601 (2008)

2008
[26]

Palmieri and D

B. Palmieri and D. Ronis, Jarzynski equality: Connec- tions to thermodynamics and the second law, Phys. Rev. E75, 011133 (2007)

2007
[27]

A. M. Monge, M. Manosas, and F. Ritort, Experimental test of ensemble inequivalence and the fluctuation theo- rem in the force ensemble in DNA pulling experiments, Phys. Rev. E98, 032146 (2018)

2018
[28]

Sone, Y.-X

A. Sone, Y.-X. Liu, and P. Cappellaro, Quantum Jarzyn- ski equality in open quantum systems from the one- time measurement scheme, Phys. Rev. Lett.125, 060602 (2020)

2020
[29]

Jarzynski, Equilibrium free-energy differences from nonequilibrium measurements: A master-equation ap- proach, Phys

C. Jarzynski, Equilibrium free-energy differences from nonequilibrium measurements: A master-equation ap- proach, Phys. Rev. E56, 5018 (1997)

1997
[30]

demonstrated that the JE is a mathematical identity following directly from Hamiltonian dynamics of the com- posite system (S+E). This derivation, later refined to incorporate strong coupling through the Hamiltonian of mean force (HMF)H ∗ β(XS , λ), confirmed that while the fr...
[31]

For such dynamics, the JE remains valid, provided the fric- tion kernel and noise correlations obey the fluctuation- dissipation theorem (FDT)

extended the equality to non-Markovian processes governed by the generalized Langevin equation. For such dynamics, the JE remains valid, provided the fric- tion kernel and noise correlations obey the fluctuation- dissipation theorem (FDT). Compliance with FDT en- sures that th...

2025 arXiv
[32]

Jarzynski, Nonequilibrium work theorem for a sys- tem strongly coupled to a thermal environment, J

C. Jarzynski, Nonequilibrium work theorem for a sys- tem strongly coupled to a thermal environment, J. Stat. Mech.: Theory Exp.2004(09), P09005

2004
[33]

Speck and U

T. Speck and U. Seifert, The Jarzynski relation, fluctu- ation theorems, and stochastic thermodynamics for non- Markovian processes, J. Stat. Mech.: Theory Exp.2007 (09), L09002

2007
[34]

Maruyama, F

K. Maruyama, F. Nori, and V. Vedral, Colloquium: The physics of Maxwell’s demon and information, Rev. Mod. Phys.81, 1 (2009)

2009
[35]

Serreli, C.-F

V. Serreli, C.-F. Lee, E. R. Kay, and D. A. Leigh, A molecular information ratchet, Nature445, 523 (2007)

2007
[36]

Maes, Local detailed balance, SciPost Phys

C. Maes, Local detailed balance, SciPost Phys. Lect. Notes. , 032 (2021)

2021
[37]

Sch¨ oll-Paschinger and C

E. Sch¨ oll-Paschinger and C. Dellago, A proof of Jarzyn- ski’s nonequilibrium work theorem for dynamical systems that conserve the canonical distribution, J. Chem. Phys. 125(2006)

2006
[38]

Gaspard and R

P. Gaspard and R. Kapral, Active matter, microre- versibility, and thermodynamics, Research (2020)

2020
[39]

I. A. Mart ´ ınez, G. Bisker, J. M. Horowitz, and J. M. Par- rondo, Inferring broken detailed balance in the absence of observable currents, Nat. Commun.10, 3542 (2019)

2019
[40]

F. S. Gnesotto, F. Mura, J. Gladrow, and C. P. Broed- ersz, Broken detailed balance and non-equilibrium dy- namics in living systems: a review, Rep. Prog. Phys.81, 066601 (2018)

2018
[41]

Goswami, Work fluctuation relations for a dragged Brownian particle in active bath, Physica525, 223 (2019)

K. Goswami, Work fluctuation relations for a dragged Brownian particle in active bath, Physica525, 223 (2019)

2019
[42]

Godec and D

A. Godec and D. E. Makarov, Challenges in inferring the directionality of active molecular processes from single- molecule fluorescence resonance energy transfer trajecto- ries, J. Phys. Chem. Lett.14, 49 (2022)

2022
[43]

Krishnamurthy, S

S. Krishnamurthy, S. Ghosh, D. Chatterji, R. Ganapathy, and A. Sood, A micrometre-sized heat engine operating between bacterial reservoirs, Nat. Phys.12, 1134 (2016)

2016
[44]

Reimann, Brownian motors: noisy transport far from equilibrium, Phys

P. Reimann, Brownian motors: noisy transport far from equilibrium, Phys. Rep.361, 57 (2002)

2002
[45]

Bechinger, R

C. Bechinger, R. Di Leonardo, H. L¨ owen, C. Reichhardt, G. Volpe, and G. Volpe, Active particles in complex and crowded environments, Rev. Mod. Phys.88, 045006 (2016)

2016
[46]

Caprini, U

L. Caprini, U. Marini Bettolo Marconi, and A. Puglisi, Activity induced delocalization and freezing in self- propelled systems, Sci. Rep.9, 1386 (2019)

2019
[47]

Vicsek and A

T. Vicsek and A. Zafeiris, Collective motion, Phys. Rep. 517, 71 (2012)

2012
[48]

M. E. Cates and J. Tailleur, Motility-induced phase sepa- ration, Annu. Rev. Condens. Matter Phys.6, 219 (2015). 7

2015
[51]

Rahbar and C

M. Rahbar and C. J. Stein, Exact fluctuation relation for open systems beyond the Zwanzig FEP equation, (2025), joint submission with this article

2025
[52]

Rahbar and C

M. Rahbar and C. J. Stein, Thermodynamic potentials from a probabilistic view on the system-environment interaction energy, arXiv preprint arXiv:2505.00188 (2025)

2025 arXiv
[53]

Talkner and P

P. Talkner and P. H¨ anggi, Colloquium: Statistical me- chanics and thermodynamics at strong coupling: Quan- tum and classical, Rev. Mod. Phys.92, 041002 (2020)

2020
[54]

M. K. Johnsrud and R. Golestanian, Fluctuation dissi- pation relations for active field theories, Phys. Rev. Res. 7, L032053 (2025)

2025
[55]

Bebon, J

R. Bebon, J. F. Robinson, and T. Speck, Thermodynam- ics of active matter: Tracking dissipation across scales, Phys. Rev. X15, 021050 (2025)

2025
[56]

M. T. Woodside and S. M. Block, Reconstructing folding energy landscapes by single-molecule force spectroscopy, Annu. Rev. Biophys.43, 19 (2014)

2014
[57]

M. P. Leighton and D. A. Sivak, Flow of energy and infor- mation in molecular machines, Annu. Rev. Phys. Chem 76(2025)

2025
[58]

Beyer and W

K. Beyer and W. T. Strunz, Operational work fluctua- tion theorem for open quantum systems, Phys. Rev. Lett. 134, 140403 (2025)

2025
[59]

Schmidt, C

L. Schmidt, C. J. Wilson, S. Behera, and B. L. de Groot, Free energy calculations for protein design, chemRxiv (2025)

2025
[60]

Kuang, K

Z. Kuang, K. M. Singh, D. J. Oliver, P. B. Dennis, C. C. Perry, and R. R. Naik, Gamma estimator of jarzynski equality for recovering binding energies from noisy dy- namic data sets, Nat. Commun.11, 5517 (2020)

2020
[61]

Casert and S

C. Casert and S. Whitelam, Learning protocols for the fast and efficient control of active matter, Nat. Commun. 15, 9128 (2024)

2024
[62]

M. C. Engel, J. A. Smith, and M. P. Brenner, Optimal control of nonequilibrium systems through automatic dif- ferentiation, Phys. Rev. X13, 041032 (2023)

2023
[63]

H. M. Baumann, V. Gapsys, B. L. de Groot, and D. L. Mobley, Challenges encountered applying equilibrium and nonequilibrium binding free energy calculations, J. Phys. Chem. B125, 4241 (2021)

2021
[64]

Serra, A

E. Serra, A. Ghidini, S. Decherchi, and A. Cavalli, Nonequilibrium binding free energy simulations: Mini- mizing dissipation, J. Chem. Theory Comput.21, 2079 (2025)
[65]

Serra, A

E. Serra, A. Ghidini, R. Aguti, M. Bernetti, S. Decherchi, and A. Cavalli, Path-based nonequilibrium binding free energy estimation, from protein–ligand to RNA-ligand binding, J. Chem. Inf. Model (2025)

2025
[66]

L. M. Cangemi, C. Bhadra, and A. Levy, Quantum en- gines and refrigerators, Phys. Rep.1087, 1 (2024)

2024
[67]

Kieninger and B

S. Kieninger and B. G. Keller, GROMACS stochastic dy- namics and BAOAB are equivalent configurational sam- pling algorithms, J. Chem. Theory Comput.18, 5792 (2022)

2022
[68]

Whitelam, Improving noisy free-energy measurements by adding more noise, Phys

S. Whitelam, Improving noisy free-energy measurements by adding more noise, Phys. Rev. E112, 014133 (2025)

2025
[69]

H¨ anggi, P

P. H¨ anggi, P. Talkner, and M. Borkovec, Reaction-rate theory: fifty years after Kramers, Rev. Mod. Phys.62, 251 (1990)

1990
[70]

Li and Z

G. Li and Z. Tu, Equilibrium free-energy differences from a linear nonequilibrium equality, Phys. Rev. E103, 032146 (2021)

2021
[71]

Saito, Relaxation in a bistable system, J

Y. Saito, Relaxation in a bistable system, J. Phys. Soc. Jpn.41, 388 (1976)

1976
[72]

S. X. Sun, Equilibrium free energies from path sampling of nonequilibrium trajectories, J. Chem. Phys.118, 5769 (2003)

2003
[73]

Zwanzig,Nonequilibrium statistical mechanics(Ox- ford university press, 2001)

R. Zwanzig,Nonequilibrium statistical mechanics(Ox- ford university press, 2001). 8 End Matter Appendix:We collect detailed proofs of the theorems and corollaries stated in the main text. Each result is re- produced in full mathematical form to ensure complete- ness and clarit...

2001

Pith tools

Reviewed August 3, 2026 · model on record in the stance chip above.

[1] [50]

Argun, A.-R

A. Argun, A.-R. Moradi, E. Pin¸ ce, G. B. Bagci, A. Im- parato, and G. Volpe, Non-Boltzmann stationary dis- tributions and nonequilibrium relations in active baths, Phys. Rev. E.94, 062150 (2016)

2016

[2] [49]

Fodor, C

´E. Fodor, C. Nardini, M. E. Cates, J. Tailleur, P. Visco, and F. Van Wijland, How far from equilibrium is active matter?, Phys. Rev. Lett.117, 038103 (2016)

2016

[3] [1]

Jarzynski, Nonequilibrium equality for free energy dif- ferences, Phys

C. Jarzynski, Nonequilibrium equality for free energy dif- ferences, Phys. Rev. Lett.78, 2690 (1997)

1997

[4] [2]

Esposito, U

M. Esposito, U. Harbola, and S. Mukamel, Nonequilib- rium fluctuations, fluctuation theorems, and counting statistics in quantum systems, Rev. Mod. Phys.81, 1665 (2009)

2009

[5] [3]

G. E. Crooks, Nonequilibrium measurements of free en- ergy differences for microscopically reversible Markovian systems, J. Stat. Phys.90, 1481 (1998)

1998

[6] [4]

G. E. Crooks, Entropy production fluctuation theorem and the nonequilibrium work relation for free energy dif- ferences, Phys. Rev. E60, 2721 (1999)

1999

[7] [5]

Liphardt, S

J. Liphardt, S. Dumont, S. B. Smith, I. Tinoco Jr, and C. Bustamante, Equilibrium information from nonequi- librium measurements in an experimental test of Jarzyn- ski’s equality, Science296, 1832 (2002)

2002

[8] [6]

Ohzeki, Quantum annealing with the Jarzynski equal- ity, Phys

M. Ohzeki, Quantum annealing with the Jarzynski equal- ity, Phys. Rev. Lett.105, 050401 (2010)

2010