REVIEW 1 major objections 5 minor 25 references
Physical Reduced Stochastic Equations for Continuously Monitored Non-Markovian Quantum Systems with a Markovian Embedding
T0 review · 1 major / 5 minor · reviewed 2026-08-07 · deepseek-v4-flash
Pith's one-line read A continuously monitored non-Markovian quantum system can be described by a closed stochastic integro-differential equation for the principal state alone, with a two-time stochastic kernel encoding the auxiliary memory.
desk verdict Sound and useful block-form stochastic Nakajima-Zwanzig equation; the abstract projection statement has a fixable right-module gap. read the letter →
The pith
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
The reading
What carries the argument
The machinery is the decomposition of the joint conditional density matrix into diagonal blocks $\varrho^{jj}_{s}$ (auxiliary labels equal) and off-diagonal blocks $\varrho^{jk}_{s}$ ($j\neq k$), through a projection superoperator $P$ and its complement $Q$. Because the probe couples through $L_0(t)$ acting only on the principal system, the stochastic back-action term $G(\rho)=L_0\rho+\rho L_0^\dagger$ commutes with $P$ and $Q$, so the off-diagonal SDE is linear in the off-diagonal blocks for any fixed diagonal-block trajectory. A stochastic variation-of-constants formula [23] solves that linear SDE with an invertible stochastic exponential $\Phi_{t,t_0}$, and substituting the solution into the diagonal-block SDE produces the two-time stochastic kernel $K_s(t,t')=L^{pq}(t)\Phi_{t,t_0}\Phi^{-1}_{t',t_0}L^{qp}(t')$, which carries the auxiliary memory. The same pattern in block form yields the kernel $A_{11}(t)\Psi_{t,t_0}\Psi^{-1}_{t',t_0}A_{00}(t')$ in Theorem 4.
What would settle it
Take any finite-dimensional projection satisfying the stated left-module property but not the right-module property, and evaluate $[P,G](\rho)$ for $G(\rho)=L_0\rho+\rho L_0^\dagger$; if it is nonzero for some $\rho$, the general version of Theorem 2 does not follow from the listed assumptions, while the explicit block-form theorem is untouched because its projection satisfies both properties. Alternatively, numerically simulate the full stochastic master equation (1) for a nontrivial auxiliary and compare its diagonal blocks with the solution of (16); any disagreement would refute the reduction.
Extended reading notes
Core claim
The central claim is that the reduced conditional state of the principal system satisfies a closed stochastic integro-differential SDE in which the off-diagonal blocks of the joint principal-auxiliary conditional density are eliminated up to their initial values. In the abstract projection formulation, Theorem 2 gives $d\varrho^p_{sa,t} = \big(L^{pp}\varrho^p_{sa,t} + L^{pq}\Phi_{t,t_0}\varrho^q_{sa,t_0} + \int_{t_0}^{t} K_s(t,t')\varrho^p_{sa,t'}dt'\big)dt + \big(G\varrho^p_{sa,t} - \varrho^p_{sa,t}\mathrm{Tr}((L_0+L_0^\dagger)\varrho^p_{sa,t})\big)dI_t$, with kernel $K_s(t,t') = L^{pq}(t)\Phi_{t,t_0}\Phi^{-1}_{t',t_0}L^{qp}(t')$, where $\Phi_{t,t_0}$ is the invertible stochastic exponential solving the linear SDE for the off-diagonal part; Theorem 4 states the same block elimination in the explicit block form with kernel $A_{11}(t)\Psi_{t,t_0}\Psi^{-1}_{t',t_0}A_{00}(t')$. Under product initial states, or under sufficiently fast asymptotic stability of the stochastic exponential, the initial off-diagonal term disappears. The diagonal blocks sum to the reduced conditional state of the principal, and their expectation gives the unconditional state, so the equation is the natural stochastic Nakajima-Zwanzig equation for a continuously monitored non-Markovian quantum system. Because the derivation starts from the homodyne measurement SDE, this equation, unlike pure-state non-Markovian quantum state diffusion, is associated with a physical continuous-measurement process.
Load-bearing premise
The load-bearing premise is that the measurement back-action commutes with the coarse-graining map $P$: the abstract proof needs $P$ to satisfy both $P((O\otimes I)\rho)=O P\rho$ and $P(\rho(O\otimes I))=(P\rho)(O\otimes I)$, but only the first is stated, while the second is used without proof. The explicit block-diagonal projection in Theorem 4 satisfies both, so this gap concerns the general statement rather than the block-form result.
Editorial extensions
If this is right
- The reduced conditional state of the principal alone obeys a closed SDE, so filtering and feedback for non-Markovian systems can work directly with the projected state, whose dimension is $n_s^2 n_a$ rather than $(n_s n_a)^2$.
- For product initial states, or any initial state diagonal in the auxiliary basis, the initial off-diagonal term vanishes, so the filter requires no knowledge of initial principal-auxiliary correlations.
- In the asymptotically stable remote-past limit, the SDE becomes a memory-only equation with integration from $-\infty$, independent of initial auxiliary data.
- The two-time stochastic kernel $K_s$ is the single object summarizing the auxiliary's influence, and the paper identifies approximating it from continuous-measurement data and using it for feedback control as the natural next problem.
- Because the equation describes the actual measurement back action, it supplies a physically interpreted filter, unlike pure-state non-Markovian quantum state diffusion, which is a computational device without a known continuous-measurement process.
Reading between the lines
- Beyond the paper, the same block elimination should apply to phase-quadrature homodyne detection and photon counting, since the paper notes the derivation is straightforwardly modified; the only structural requirement is that the measurement operator act on the principal system alone.
- Beyond the paper, the right-module gap in the abstract projection argument suggests that any use of Theorem 2 outside the explicit block projection should first verify both commutation properties; the block-form Theorem 4 is not affected because its projection satisfies them.
- Beyond the paper, if the two-time kernel can be estimated from experimental records, one could implement a data-driven non-Markovian quantum filter and controller without ever modelling the auxiliary system.
- Beyond the paper, averaging the new stochastic equation should reproduce the Nakajima-Zwanzig master equation; this means the practical payoff is more likely in real-time estimation and feedback than in Monte Carlo computation of unconditional states, where the projected density matrix has no size advantage over the deterministic equation.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper derives closed stochastic integro-differential equations for the reduced conditional state of a principal system in a Markovian embedding of a non-Markovian quantum system under continuous homodyne measurement. Building on the coupled principal-operator SDE system of [12], the author introduces an abstract projection P onto the 'diagonal' blocks in an auxiliary basis, splits the coupled system into p = P and q = Q components, and eliminates the q-component via a stochastic variation-of-constants formula (Lemma 1, citing [23]). This yields a closed SDE for the p-component alone (Theorem 2, Eq. (7)) with a two-time stochastic kernel K_s(t,t') = L(t)^{pq} Phi_{t,t0} Phi^{-1}_{t',t0} L(t')^{qp}; the initial-condition term drops out under product initial states or a suitable asymptotic-stability limit. The same procedure is then carried out explicitly in block form (Eqs. (10)-(17), Lemma 3, Theorem 4), giving the kernel K_s(t,t') = A11(t) Psi_{t,t0} Psi^{-1}_{t',t0} A00(t'). The result is presented as a stochastic Nakajima-Zwanzig-type equation, and the paper is careful to state where the derivation does not provide computational advantage for unconditional states and to defer stability conditions to the cited literature.
Significance. If the reduction is valid, this is a useful and conceptually clean result: it provides the conditional reduced dynamics of the embedded non-Markovian system with a concrete continuous-measurement interpretation and a closed-form memory-kernel structure, complementing NMQSD-type approaches. The derivation is algebraic and parameter-free, with no fitted quantities; the explicit block form is internally consistent, and the structural correspondence between the abstract Theorem 2 and the block-form Theorem 4 is transparent. The paper also deserves credit for honestly flagging that Monte Carlo simulation of the reduced equation is not computationally more efficient than solving the Nakajima-Zwanzig equation for unconditional states, and for deferring asymptotic-stability analysis to explicit references rather than overclaiming. The principal shortcoming, a missing hypothesis in the abstract projection formulation, is local and fixable and does not endanger the concrete block-form result.
major comments (1)
- [Section III (properties of P, sentence before Eq. (3))] The sentence 'by the assumptions on P, G(t) commutes with both P and Q' does not follow from the two stated properties of P. Property 2 is a left-module identity, P((O tensor I)rho) = O P rho. To pass P through the second term of G(t)rho = L0(t)rho + rho L0(t)^dagger one needs the right-module identity P(rho (O tensor I)) = (P rho)(O tensor I), which is used when the projector acts on the term rho L0(t)^dagger but is neither stated nor proved. Consequently Eqs. (3) and (4), and hence Theorem 2 as stated for a general P, are not derived from the listed hypotheses. The explicit block-diagonal projection P X = sum_j X^{jj} tensor |phi_j><phi_j| does satisfy both module identities by direct verification, so the block-form Theorem 4 and its kernel are not endangered. The fix is to add the right-module identity as an explicit assumption on P (noting that both the conditional-expectation example P rho = Tr_a(rho) tensor rho_a and the block-diagonal projection satisfy it), or to state Theorem 2 only for bimodule projections.
minor comments (5)
- [Section III (before Lemma 3)] The text states that the proofs of Lemma 3 and Theorem 4 'follow by the same proof steps as for Lemma 1 and Theorem 2' and that they are omitted; this is an acknowledged proof omission for the explicit block-form results, and it should be flagged in the published record. I verified that the substitutions A01(t)dt + B01(t)dI maps onto dA(t) and A00(t)dt maps onto L(t)^{qp}dt, so Eq. (13) has the same structure as Eq. (5), and Eq. (14) matches Eq. (3); the mirroring is structurally sound, but the two proofs are short and should be included in the arXiv or journal version, or the term-by-term identification should be stated explicitly.
- [Conclusion (kernel formula)] In the Conclusion, the block-form kernel is printed as K_s(t,t') = A11(t) Psi_{t,t0} Psi^{-1}_{t,t0} A00(t'); the second factor should be Psi^{-1}_{t',t0}, matching Eq. (16). In the same sentence, 'in SDEs (3) and (9)' should read '(7) and (9)', since Eq. (3) does not contain the kernel.
- [Section III (after Eq. (5))] The invertibility of the stochastic exponential Phi (and likewise Psi) is asserted with the single citation [23]; please quote the specific theorem and its hypotheses, such as continuity or boundedness of the coefficients on finite time intervals, so that a reader can verify that the present coefficients, which depend on the random process rho^p, satisfy them pathwise.
- [Section III (Theorems 2 and 4)] Theorems 2 and 4 claim the forward implication that solutions of the original filter (1) satisfy the reduced equations; the well-posedness of (7) and (16) as standalone path-dependent integro-differential SDEs is not discussed. Since the kernel K_s depends on the path of rho^p through Phi (or Psi), a clarifying sentence on the status of the converse, or an explicit remark that well-posedness of the reduced equation is open, would help the reader.
- [Throughout] Minor typographical issues: 'supeoperator' in Section III; 'equatiom' in the discussion of the Nakajima-Zwanzig equation; 'is is based' in the Introduction paragraph on NMQSD; and 'filed' in the title of reference [7]. These do not affect the mathematics.
Circularity Check
No significant circularity: Theorem 2 is a genuine variation-of-constants elimination with an emergent kernel, and the reliance on the author's prior [12] is limited to the input block SDEs.
full rationale
The central derivation (Theorem 2, Eq. (7)) is obtained by substituting the explicit solution of the q-component SDE (Lemma 1) into the p-component SDE (3). The solution of the linear SDE (5) uses the external matrix-valued stochastic variation-of-constants theorem of Duan and Yan ([23]), not a result of the present paper. The two-time stochastic kernel Ks(t,t') = L(t)^pq Phi_{t,t0} Phi^{-1}_{t',t0} L(t')^qp is not an input or a fitted quantity; it emerges from the substitution and is built from the model's superoperators. The block-form Theorem 4 follows by 'the same proof steps' (explicitly stated after Lemma 3), so it inherits the same non-circular structure. The paper's only self-citation is to the author's prior work [12], which supplies the starting point: the coupled block SDEs (10) for the principal-system operators. That prior derivation is an independent, published derivation (CDC 2023) of the joint filter projected onto auxiliary basis blocks, and it is not assumed to contain the target result: the closed diagonal-block SDE with the kernel Ks does not appear in [12] and is derived here for the first time. No parameter is fitted to data and no equation equivalent to the conclusion is inserted as an assumption. The reviewer-identified gap -- that Section III states only the left-module property P((O otimes I)rho) = O P rho and then asserts without proof that G(t) commutes with P, which also requires the right-module property P(rho(O otimes I)) = (P rho)(O otimes I) -- is a completeness/correctness issue in the general abstract Theorem 2, not a circularity: the missing right-module property is not the conclusion of the theorem, and the explicit block-diagonal projection used in Theorem 4 is directly verified to satisfy both properties. Accordingly, the circularity score is low (1), reflecting one self-citation that is not load-bearing in a circular sense.
Assumptions & free parameters
assumptions (6)
- standard math Standard Itô stochastic calculus and the Duan-Yan variation-of-constants theorem for matrix-valued linear SDEs.
- standard math The stochastic exponential Φ (and Ψ) is pathwise invertible.
- domain assumption Existence and pathwise uniqueness of the joint conditional-state SDE (1) under Assumption 5.1 of [22].
- domain assumption The auxiliary Hilbert space h_a is finite dimensional and all superoperators can be represented by finite-dimensional real matrices.
- domain assumption The measurement probe couples to the principal system only, L0(t) ∈ L(hs).
- ad hoc to paper The projection P additionally acts as a right-module map over principal-system operators (used, not stated).
Cite this review
Pith. "Pith review of Physical Reduced Stochastic Equations for Continuously Monitored Non-Markovian Quantum Systems with a Markovian Embedding." pith.science (2026). https://pith.science/paper/URG2WWSE
@misc{pith2026250522070,
author = {Pith},
title = {Pith review of: Physical Reduced Stochastic Equations for Continuously Monitored Non-Markovian Quantum Systems with a Markovian Embedding},
year = {2026},
howpublished = {\url{https://pith.science/paper/URG2WWSE}},
note = {Machine review of arXiv:2505.22070}
}
read the original abstract
An effective approach to modeling non-Markovian quantum systems is to embed a principal (quantum) system of interest into a larger quantum system. A widely employed embedding is one that uses another quantum system, referred to as the auxiliary system, which is coupled to the principal system, and both the principal and auxiliary can be coupled to quantum white noise processes. The principal and auxiliary together form a quantum Markov system and the quantum white noises act as a bath (environment) for this system. Recently it was shown that the conditional evolution of the principal system in this embedding under continuous monitoring by a travelling quantum probe can be expressed as a system of coupled stochastic differential equations (SDEs) that involve only operators of the principal system. The reduced conditional state of the principal only (conditioned on the measurement outcomes) is determined by the ``diagonal" blocks of this coupled systems of SDEs. It is shown here that the ``off-diagonal" blocks can be exactly eliminated up to their initial conditions, leaving a reduced closed system of SDEs for the diagonal blocks only. Under additional conditions the off-diagonal initial conditions can be made to vanish. This new closed system of equations, which includes an integration term involving a two-time stochastic kernel, represents the non-Markovian stochastic dynamics of the principal system under continuous-measurement. The system of equations determine the reduced conditional state of the principal only and may be viewed as a stochastic Nakajima-Zwanzig type of equation for continuously monitored non-Markovian quantum systems.
Reference graph
Works this paper leans on
-
[12]
H. I. Nurdin, “Markovian embeddings of non-Markovian quantum systems: Coupled stochastic and quantum master equations for non- Markovian quantum systems,” in Proc. IEEE Conference on Decision and Control, pp. 5570 - 5576 , 2023, pp. 5570–5576
work page 2023
-
[23]
General matrix-valued inhomogeneous linear stochastic differential equations and applications,
J. Duan and J. Yan, “General matrix-valued inhomogeneous linear stochastic differential equations and applications,” Stat. Prob. Lett. , vol. 78, no. 15, pp. 2361–2365, 2008
work page 2008
-
[1]
Quantum Ito’s formula and stochastic evolution,
R. L. Hudson and K. R. Parthasarathy, “Quantum Ito’s formula and stochastic evolution,” Commun. Math. Phys. , vol. 93, pp. 301–323, 1984
work page 1984
-
[2]
C. W. Gardiner and P. Zoller, Quantum Noise: A Handbook of Markovian and Non-Markovian Quantum Stochastic Methods with Applications to Quantum Optics , 3rd ed. Berlin and New York: Springer-Verlag, 2004
work page 2004
-
[3]
An introduction to quantum filtering,
L. Bouten, R. van Handel, and M. R. James, “An introduction to quantum filtering,” SIAM J. Control Optim. , vol. 46, pp. 2199–2241, 2007
work page 2007
-
[4]
H. M. Wiseman and G. J. Milburn, Quantum Measurement and Control. Cambridge University Press, 2010
2010
-
[5]
H. I. Nurdin and N. Yamamoto, Linear Dynamical Quantum Systems: Analysis, Synthesis, and Control , ser. Communications and Control Engineering. Cham: Switzerland: Springer, 2017
work page 2017
-
[6]
M. Nielsen and I. Chuang, Quantum Computation and Quantum Information: 10th Anniversary Edition . Cambridge University Press, 2010
work page 2010
Show all 25 references
-
[7]
Quantum master equation and filter for systems driven by filed in a single photon state,
J. Gough, M. R. James, and H. I. Nurdin, “Quantum master equation and filter for systems driven by filed in a single photon state,” in Proc. IEEE Conference on Decision and Control , 2011, pp. 5570–5576
2011
-
[8]
Quantum filter- ing for systems driven by fields in single-photon states or superposition of coherent states,
J. Gough, M. R. James, H. I. Nurdin, and J. Combes, “Quantum filter- ing for systems driven by fields in single-photon states or superposition of coherent states,” Phys. Rev. A , vol. 86, p. 043819, 2012
2012
-
[9]
Quantum filtering for systems driven by fields in single photon states and superposition of coherent states using non-markovian embeddings,
J. E. Gough, M. R. James, and H. I. Nurdin, “Quantum filtering for systems driven by fields in single photon states and superposition of coherent states using non-markovian embeddings,” Quantum Inf Process, vol. 12, pp. 1469–1499, 2013
2013
-
[10]
Quantum trajectories for a class of continuous matrix product input states,
——, “Quantum trajectories for a class of continuous matrix product input states,” New J. Phys. , vol. 16, p. 075008, 2014
2014
-
[11]
Introduction to quantum noise, measurement, and amplification,
A. A. Clerk, M. H. Devoret, S. M. Girvin, F. Marquardt, and R. J. Schoelkopf, “Introduction to quantum noise, measurement, and amplification,” Rev. Mod. Phys. , vol. 82, p. 1155, 2010
2010
-
[13]
Stochastic wave-function approach to non-Markovian systems,
A. Imamoglu, “Stochastic wave-function approach to non-Markovian systems,” Phys. Rev. A , vol. 50, no. 5, pp. 3650–3653, 1994
1994
-
[14]
Theory of pseudomodes in quantum optical processes,
B. J. Dalton, S. M. Barnett, and B. M. Garraway, “Theory of pseudomodes in quantum optical processes,” Phys. Rev. A , vol. 64, p. 053813, 2001
2001
-
[15]
Optimized auxiliary oscillators for the simulation of general open quantum systems,
F. Mascherpa et al., “Optimized auxiliary oscillators for the simulation of general open quantum systems,” Phys. Rev. A , vol. 101, p. 052108, 2020
2020
-
[16]
Markovian treatment of non-Markovian dynamics of open fermionic systems,
F. Chen, E. Arrigoni, and M. Galperin, “Markovian treatment of non-Markovian dynamics of open fermionic systems,” New J. Phys. , vol. 21, p. 123035, 2019
2019
-
[17]
The SLH framework for modeling quantum input-output networks,
J. Combes, J. Kerckhoff, and M. Sarovar, “The SLH framework for modeling quantum input-output networks,” Adv. Phys. X , vol. 2, no. 784, 2017
2017
-
[18]
Carmichael, An Open Systems Approach to Quantum Optics
H. Carmichael, An Open Systems Approach to Quantum Optics . Berlin: Springer, 1993
1993
-
[19]
Non-Markovian quantum state diffusion,
L. Di ´osi, N. Gisin, and W. T. Strunz, “Non-Markovian quantum state diffusion,” Phys. Rev. A , vol. 58, p. 1699, 1998
1998
-
[20]
Non-Markovian quantum trajectories, instruments and time-continuous measurements,
S. Kr ¨onke and W. T. Strunz, “Non-Markovian quantum trajectories, instruments and time-continuous measurements,” J. Phys. A , vol. 45, no. 5, p. 055305, 2012
2012
-
[21]
Reproducing kernel Hilbert space approach to non-Markovian quantum stochastic models,
J. Gough, H. Ding, and N. H. Amini, “Reproducing kernel Hilbert space approach to non-Markovian quantum stochastic models,” J. Math. Phys. , p. 042102, 2025
2025
-
[22]
The stochastic master equation: Part II,
A. Barchielli and M. Gregoratti, “The stochastic master equation: Part II,” in Quantum Trajectories and Measurements in Continuous Time , ser. Lecture Notes in Physics, 2009, vol. 782, pp. 111–123
2009
-
[24]
Non-Markovian dynamics in open quantum systems,
B. Vacchini, “Non-Markovian dynamics in open quantum systems,” 2011, lecture notes 2224-2 in the “School on New Trends in Quantum Dynamics and Quantum Entanglement” (14-18 February 2011), The Abdussalam Centre for Theoretical Physics
2011
-
[25]
On asymptotic behavior of solutions of linear multidimensional stochastic differential equations with multiplicative noise,
E. S. Palamarchuk, “On asymptotic behavior of solutions of linear multidimensional stochastic differential equations with multiplicative noise,” Theory Probab. Appl. , vol. 69, no. 3, pp. 372–390, 2024
2024
Reviewed August 7, 2026 · model on record in the stance chip above.
Discussion (0). Sign in to comment.