Pith. sign in

REVIEW 2 major objections 4 minor 1 cited by

The Thermodynamic Cost of Ignorance: Thermal State Preparation with One Ancilla Qubit

T0 review · 2 major / 4 minor · reviewed 2026-08-09 · deepseek-v4-flash

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

Pith's one-line read The paper proves that repeated random couplings of one thermalized ancilla qubit drive any non-degenerate quantum system to its thermal state, and bounds the total simulation time, quantifying the cost of ignorance about the spectrum.

desk verdict Nice single-ancilla thermalization idea, but Theorem 9's remainder bound is wrong by dimension powers, so the runtime theorems collapse as written; still worth refereeing. read the letter →

arxiv 2502.03410 v1 pith:M5VXQZ3Z submitted 2025-02-05 quant-ph

Matthew Hagan , Nathan Wiebe This is my paper · ORCID
classification quant-ph MSC 81P6881S2260J10
keywords thermalstatepreparationGibbsweak-couplingexpansionMarkovchainspectralgaprepeatedinteractionssingle-ancillaprotocolopenquantumsystemsHamiltoniansimulation
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

The paper claims that a simple physical process can be turned into a working quantum algorithm: prepare one ancilla qubit in a thermal state, couple it to the system via a random interaction, let the pair evolve, discard the qubit, and repeat. In the weak-coupling limit the authors prove these dynamics reduce to a classical Markov chain over the system's energy eigenstates whose fixed point is the thermal state (at finite $\beta$) or the ground state (as $\beta\to\infty$), which lets them bound the total simulation time in terms of the chain's spectral gap $\lambda_*(\beta)$. They prove two runtime theorems: with no knowledge of the spectrum, total time is $\widetilde{O}(\dim_S^{16}\|H_S\|^7/(\delta_{\min}^8\epsilon^6\lambda_*(\beta)^7))$ (Theorem 12); with perfect knowledge of eigenvalue differences it drops to $\widetilde{O}(\dim_S^{16}/(\delta_{\min}\epsilon^{2.5}\lambda_*(\beta)^{3.5}))$ (Theorem 13), a ratio that quantifies the 'thermodynamic cost of ignorance.' If the claims hold, this gives quantum computers a thermal-state preparation subroutine that uses exactly one ancilla qubit, compiles to time-independent Hamiltonian simulation with no rejection or filtering steps, and extends the repeated-interactions picture of open quantum systems to arbitrary non-degenerate Hamiltonians.

What carries the argument

The central object is the randomized interaction $G=U_{\rm Haar}DU_{\rm Haar}^\dagger$ with Gaussian eigenvalues, whose Haar average can be evaluated in closed form and makes the second-order expansion of the channel exactly computable. The analysis then reduces the quantum channel to a column-stochastic matrix $I+T_{\rm on}$ acting on the diagonal of the density matrix, with transition amplitudes $\tilde\alpha^2\,\mathrm{sinc}^2((\Delta_S(i,j)\pm\gamma)t/2)$; the thermal state is a fixed point when a detailed-balance-like sum (Eq. 63) vanishes, which holds exactly when $\gamma$ is sampled from the true eigenvalue differences and only approximately under uniform guessing. The remaining work is the Markov-chain spectral gap $\lambda_*(\beta)$: it converts into an interaction count through a Markov relaxation bound, and the paper computes it exactly in the ground-state limit, where the transition matrix becomes upper triangular and the gap is a constant fixed by the spectrum's difference multiplicities.

What would settle it

Numerically evaluate the third-order remainder for a small system: for a two- or three-qubit non-degenerate $H_S$ plus one ancilla qubit, approximate the Haar/Gaussian average with many random draws of $G$, exactly exponentiate $H+\alpha G$ at moderate $\alpha t$, and compare the channel output with the second-order expansion to test whether the trace-norm remainder obeys $\|R_\Phi(\rho)\|_1\le 16\sqrt{2/\pi}\,\dim_S(\alpha t)^3$ or grows further with dimension. A cheaper, calculation-only check: sample i.i.d. Gaussian $d_i$ in dimension $d$ and compare $\mathbb{E}(\sum_i |d_i|)^3$ with the value $2d\,\mathbb{E}|y|^3$ used in the proof; the cross terms make the two differ for $d>1$.

Watch

Extended reading notes

Core claim

For a non-degenerate system Hamiltonian $H_S$ and a single-qubit environment with gap $\gamma$, the paper studies the channel $\Phi(\rho)=\mathrm{Tr}_{\rm Env}\mathbb{E}_G[e^{-i(H+\alpha G)t}(\rho\otimes e^{-\beta H_E}/Z)e^{i(H+\alpha G)t}]$, where $G=U_{\rm Haar}DU_{\rm Haar}^\dagger$ has Haar-random eigenvectors and i.i.d. Gaussian eigenvalues. Expanding to second order in the coupling $\alpha$, the authors show coherences vanish and the channel acts as a classical Markov chain on the eigenstates of $H_S$: transitions $|i\rangle\to|j\rangle$ are set by sinc-squared resonance factors $\mathrm{sinc}^2((\Delta_S(i,j)\pm\gamma)t/2)$, while the off-resonance part contributes an error bounded by $8\alpha^2/\delta_{\min}^2$ and the truncation remainder by $16\sqrt{2/\pi}\,\dim_S(\alpha t)^3$ (Theorem 9). For a uniformly random $\gamma\in[0,4\|H_S\|]$ the thermal state is an approximate fixed point with explicit deviation (Eq. 151), and in the $\beta\to\infty$ limit it is the exact fixed point with the spectral gap of the rescaled transition matrix lower bounded by a constant. Feeding the gap into the paper's Markov relaxation bound (Theorem 6) yields total simulation time $\widetilde{O}(\dim_S^{16}\|H_S\|^7/(\delta_{\min}^8\epsilon^6\lambda_*(\beta)^7))$ under zero knowledge (Theorem 12) and $\widetilde{O}(\dim_S^{16}/(\delta_{\min}\epsilon^{2.5}\lambda_*(\beta)^{3.5}))$ when all eigenvalue differences are known (Theorem 13).

Load-bearing premise

The load-bearing premise is the Theorem 9 estimate, proved in the appendix: the error from stopping the expansion of the channel at second order in the coupling is at most a constant times the system dimension times $(\alpha t)^3$; every runtime bound in the paper includes this quantity in its error budget, so if the estimate is too small the stated runtimes do not follow.

Editorial extensions

If this is right

  • A quantum computer could prepare the Gibbs state of any non-degenerate Hamiltonian at inverse temperature $\beta$ with a single ancilla qubit, time-independent Hamiltonian simulation, and no rejection, filtering, or Lindbladian jump-operator engineering.
  • Spectral knowledge becomes a rigorously quantifiable resource: knowing all eigenvalue differences improves the proven total simulation time by a factor $\widetilde{O}(\|H_S\|^7/(\delta_{\min}^7\epsilon^{3.5}\lambda_*(\beta)^{3.5}))$, the first explicit worst-case accounting of eigenvalue heuristics in thermal state preparation.
  • In the $\beta\to\infty$ limit the same channel prepares the ground state, since the ground state is the unique fixed point and the spectral gap of the rescaled transition matrix is a computable constant; this answers two open questions posed by the prior state of the art, namely thermalizing all non-degenerate Hamiltonians and obtaining a computed gap in the ground-state limit.
  • The repeated-interactions model of open quantum systems is extended from specific system-environment pairs to arbitrary non-degenerate Hamiltonians with completely unknown interactions, giving a picture of thermalization for finite resources instead of thermodynamic limits.
  • The paper's numerics indicate the proven bounds are loose: the channel still converges at couplings far beyond the analyzed weak-coupling regime, and empirically the total time scales as roughly $\epsilon^{-0.68}$ for $\alpha\propto 1/t$ in the harmonic oscillator, nearly four orders of magnitude below the proven $\epsilon^{-2.5}$ bound at $\epsilon\approx 0.005$.

Reading between the lines

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

  • The 'cost of ignorance' is quantified only at two extremes, but the single-qubit theorem points to a continuum: a user who knows a window of width $\sigma$ containing each gap should interpolate between the two costs, and deriving that interpolation for general systems is a direct next step the paper leaves open.
  • Because the ancilla's transition statistics encode the system's energy gaps and temperature, this channel could double as an interaction-agnostic thermometer or spectrometer; the paper only gestures at this, so bounding the estimation error (for instance via the Fisher information of the ancilla output) would be a concrete extension.
  • The remainder estimate flagged in the weakest-assumption field is checkable by a small numerical experiment (see the falsifier): if the missing cross-term contribution to $\mathbb{E}\|G\|_1^3$ is modest, the theorems survive with adjusted constants, whereas a large contribution would shrink the valid weak-coupling regime that the paper's parameter choices rely on.
  • Since the channel reduces to a Markov chain, the thermalization process can be classically simulated for small molecules: one could compute the finite-temperature spectral gap $\lambda_*(\beta)$ for the paper's hydrogen-chain examples and test whether the $\beta\to\infty$ gap formula remains predictive at intermediate temperature.
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.

Referee Report

2 major / 4 minor

Summary. The paper proposes a thermal state preparation protocol in which a single ancilla qubit, prepared in a thermal state, interacts with the system through a random interaction G with Haar-random eigenvectors and i.i.d. Gaussian eigenvalues. The system-ancilla evolution is repeated L times, and the channel Φ is expanded to second order in the coupling α. The authors show that, in the weak-coupling regime, the diagonal part of the dynamics reduces to a classical Markov chain whose transition matrix they compute, and they argue that the Gibbs state is an approximate fixed point. They then state resource bounds for single-qubit systems, truncated harmonic oscillators, and general non-degenerate Hamiltonians, comparing a zero-knowledge setting with a perfect-eigenvalue-knowledge setting, and they report numerical experiments on hydrogen chains.

Significance. If the main bounds were correct, this would be a valuable contribution: it offers a minimal-ancilla thermal state preparation routine with a direct Markov-chain analysis, gives an explicit detailed-balance condition rather than relying on the KMS condition, computes the spectral gap exactly in the ground-state limit, and provides the first explicit comparison between complete ignorance and perfect eigenvalue knowledge in this setting. The numerical experiments are also a useful complement to the analytic claims. However, the central quantitative claims all rest on the remainder bound in Theorem 9, and that bound appears to be miscalculated in a way that changes the dimension scaling of the runtimes.

major comments (2)
  1. [Appendix A.3 / Theorem 9, Eq. (A104)] The key equality in the proof of Theorem 9 is incorrect. The text states that ∫||G||_1^3 dG = ∫||D||_1^3 dU dD = Σ_i ∫|d_i|^3 d d_i, but ||D||_1 = Σ_i |d_i|, so ||D||_1^3 contains many positive cross terms. For n = 2 dim_S i.i.d. standard Gaussian eigenvalues, one has E||D||_1^3 = 2c n + 3c n(n−1) + c^3 n(n−1)(n−2) with c = √(2/π), which is Θ(n^3), not Θ(n). Consequently the claimed bound ||RΦ(ρ)||_1 ≤ 16√(2/π) dim_S(αt)^3 is missing a factor that grows as dim_S^2, and the correct remainder bound is of order dim_S^3(αt)^3. This remainder is used in the error decompositions of Theorems 10–13, e.g., Eqs. (88), (133), (176), and (206), so the parameter settings in Eqs. (84), (113), (152), and (193) and the total simulation time claims in Eqs. (85), (115), (153), and (194) are not established as written. The qualitative mechanism may survive, but the quantitative results require a recomputation of the error balance with the corrected dimension dependence.
  2. [Theorem 12, Eq. (191)] The proof of the lower bound lim_{β→∞} λ⋆(β) ≥ 2.43 in the zero-knowledge ground-state limit relies on the inequality chain 2 ≤ 32/π ≤ dim_S^2/π in Eq. (191). This chain is false for dim_S < 6; for example, dim_S = 3 gives 32/π ≈ 10.19 while dim_S^2/π ≈ 2.86. The statement that the sinc integral is bounded below by a constant therefore needs a revised argument, or the theorem needs an explicit dimensional restriction under which the claimed inequality is valid.
minor comments (4)
  1. [Throughout, Eq. (5)] The notation eα2 is used to denote (αt)^2/(dim+1), but as printed it resembles "e times α squared" rather than a defined symbol such as \tilde{\alpha}^2. Please introduce a clearer notation.
  2. [Theorem 12, Eq. (151) and Eq. (178)] Eq. (151) and Eq. (178) use the constant 16√(π/2) for the remainder term, while Theorem 9 states 16√(2/π). These constants do not match and should be harmonized, independently of the larger issue in the remainder bound itself.
  3. [Theorem 12, setup] The statement of Theorem 12 should specify the regime in which δ_min is well defined. Since δ_min is defined in Eq. (3) as a minimum over pairs of distinct eigenvalue differences, it is undefined for a two-level system; the paper treats the two-level case separately in Theorem 10 but should still state the dimensional assumption needed for Theorem 12.
  4. [Section III C, Fig. 3] The numerical slopes in Fig. 3 are reported to be consistently about 0.18 larger than the analytic predictions. The text notes this but does not explain whether this is a systematic artifact of the fits or an indication of looseness in the analytic bounds; a brief comment would be helpful.

Circularity Check

0 steps flagged · score 0.0 of 10
full rationale

The central derivation is self-contained in the sense required here. The channel Phi is defined (Eq. 7) with an independent input: a single thermal ancilla e^{-beta H_E}/Z; the target is the system Gibbs state e^{-beta H_S}/Z. The weak-coupling expansion (Lemma 2, Theorem 3) computes transition amplitudes from the channel without assuming the fixed point. Lemma 5 converts the diagonal dynamics to a classical Markov chain, and Lemma 7 gives a detailed-balance criterion for fixed points of I+T. Theorems 12 and 13 then compute the transition probabilities E_gamma<j|T_on^(gamma)(|i><i|)|j> from Eq. (20)/(26), insert them into the Lemma 7 criterion, and verify either approximate (Eqs. 162-175) or exact (Eqs. 200-205) detailed balance against the Gibbs weights e^{-beta lambda_S(i)}/Z_S(beta). The fixed point is therefore derived, not imposed: no parameter is fitted to the thermal state and no 'prediction' is an input renamed. The only self-citations ([32],[46]) appear in the introduction and conclusion for comparison to repeated-interaction conversion and for product-formula circuit compilation; they are not load-bearing for the fixed-point or runtime theorems. The imported Markov-chain bound (Jerison, Theorem 6) and Haar-integral identity (Lemma 15 from [47]) are external results, not author-imposed uniqueness claims. The possible error in the Theorem 9 remainder bound identified by the reviewer is a correctness and rigor concern about the error budget, not a circularity: an incorrect bound would invalidate the quantitative runtime claims, but it does not make the conclusion equivalent to an input. Limitations, such as the inability to compute finite-beta spectral gaps and the open strong-coupling analysis, are explicitly acknowledged in Sections III.C and V and do not indicate circularity.

Assumptions & free parameters 0 free parameters · 6 assumptions · 0 invented entities

The central claim rests on standard Haar measure and Markov chain tools, on the physical assumptions of a single-qubit thermal environment and weak coupling, and on the paper-specific choice of the random interaction ensemble. No constants are fitted to data. The proof of the remainder bound, however, contains a dimension-scaling error, so the ledger of what is actually established is smaller than the paper claims.

assumptions (6)
  • domain assumption The system Hamiltonian H_S has no degenerate eigenvalues.
    Assumed throughout Section IV to make the transition matrix upper triangular in the beta->infinity limit and to define delta_min; stated before Theorem 12.
  • domain assumption The input state commutes with H_S.
    The Markov chain reduction is proven only for density matrices diagonal in the eigenbasis of H_S; stated in Theorems 10-13.
  • standard math Haar integration formulas (Weingarten calculus, Lemma 15).
    Used to evaluate averages of products of random interaction eigenvectors; cited from [47].
  • standard math Jerison's Markov relaxation bound (Theorem 6).
    Used to convert spectral gaps into mixing-time bounds for non-reversible Markov chains; cited to [34].
  • domain assumption The second-order weak-coupling Taylor expansion is valid with a Taylor remainder of the stated form.
    The entire channel analysis is based on an expansion in alpha about 0; the remainder is Theorem 9, whose proof contains the suspected dimension-scaling error.
  • ad hoc to paper The random interaction G has Haar-distributed eigenvectors and i.i.d. standard Gaussian eigenvalues (Eq. 4).
    This specific model is introduced by the authors to enable Haar integration; it is not derived from physical principles but is part of the protocol definition.

how reviews work

0 comments
Cite this review

Pith. "Pith review of The Thermodynamic Cost of Ignorance: Thermal State Preparation with One Ancilla Qubit." pith.science (2026). https://pith.science/paper/M5VXQZ3Z

@misc{pith2026250203410,
  author       = {Pith},
  title        = {Pith review of: The Thermodynamic Cost of Ignorance: Thermal State Preparation with One Ancilla Qubit},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/M5VXQZ3Z}},
  note         = {Machine review of arXiv:2502.03410}
}
abstract

In this work we investigate a model of thermalization wherein a single ancillary qubit randomly interacts with the system to be thermalized. This not only sheds light on the emergence of Gibbs states in nature, but also provides a routine for preparing arbitrary thermal states on a digital quantum computer. For desired $\beta$ and random interaction $G$ the routine boils down to time independent Hamiltonian simulation and is represented by the channel $\Phi : \rho \mapsto \mathbb{E}_G {\rm Tr}_{\rm Env} \left[ e^{-i(H + \alpha G)t} \left(\rho \otimes \frac{e^{-\beta H_E}}{\mathcal{Z}}\right) e^{i (H + \alpha G)t} \right]$. We rigorously prove that these dynamics reduce to a Markov chain process in the weak-coupling regime with the thermal state as the approximate fixed point. We upper bound the total simulation time required in terms of the Markov chain spectral gap $\lambda_\star$, which we compute exactly in the ground state limit. These results are independent of any eigenvalue knowledge of the system, but we are further able to show that with knowledge of eigenvalue differences $\lambda_S(i) - \lambda_S(j)$, then the total simulation time is dramatically reduced. The ratio of the complete ignorance simulation cost to the perfect knowledge simulation cost scales as $\widetilde{O} \left({\frac{\|{H_S}\|^7}{\delta_{\rm min}^7 \epsilon^{3.5} \lambda_\star(\beta)^{3.5}}}\right)$, where $\delta_{\min}$ is related to the eigenvalue differences of the system. Additionally, we provide more specific results for single qubit and harmonic oscillator systems as well as numeric experiments with hydrogen chains. In addition to the algorithmic merits, these results can be viewed as broad extensions of the Repeated Interactions model to generic Hamiltonians with unknown interactions, giving a complete picture of the thermalization process for quantum systems.

Figures

Figures reproduced from arXiv: 2502.03410 by the authors.

Figure 1
Figure 1. FIG. 1: Total simulation time for a single qubit system to reach within trace distance of 0 [PITH_FULL_IMAGE:figures/full_fig_p024_1.png] view at source ↗
Figure 2
Figure 2. FIG. 2: Demonstration of [PITH_FULL_IMAGE:figures/full_fig_p025_2.png] view at source ↗
Figure 3
Figure 3. FIG. 3: Scaling of [PITH_FULL_IMAGE:figures/full_fig_p026_3.png] view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: FIG. 4: These plots show the distance to the target thermal state for Hydrogen 2 and Hydrogen 3 [PITH_FULL_IMAGE:figures/full_fig_p035_4.png]
Figure 5
Figure 5. Figure 5: FIG. 5: In these plots the amount of total simulation time needed to prepare a [PITH_FULL_IMAGE:figures/full_fig_p036_5.png]

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

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

  1. Collisional model with dissipative and dephasing baths: Nonadditive effects at strong coupling

    quant-ph 2025-09 conditional novelty 6.0 of 10

    An exact solution of a two-bath collision model shows nonadditive bath effects at strong coupling, including dephasing-induced slowing of relaxation and zero heat flow to the dephasing bath.

Reference graph

Works this paper leans on

54 extracted references · 46 canonical work pages · cited by 1 Pith paper

  1. [1]

    For finite β the thermal state is an approximate fixed point of the thermalizing channel EγΦγ with a deviation of ||ρS(β) − EγΦγ(ρS(β))||1 ≤ α2teβδmin ||HS||−1 π + 8 α2 δmin + 16 r π 2 dimS(αt)3. (151)

    work page

  2. [2]

    The total simulation time needed is therefore L · t ∈ eO dim16 S ||HS||7 δ8 minϵ6eλ⋆(β)7 !

    The parameter settings for any β ∈ [0, ∞] and error tolerance ϵ ∈ (0, 2] α = δ4 minϵ3eλ⋆(β)3 dim7 S ||HS||3 , t = dim2 S ||HS|| ϵeλ⋆(β)δ2 min , and L ∈ eO dim14 S ||HS||6 ϵ5δ6 mineλ⋆(β)6 ! (152) are sufficient to guarantee ρfix − (EγΦγ)◦L (ρ) 1 ∈ eO (ϵ). The total simulation time needed is therefore L · t ∈ eO dim16 S ||HS||7 δ8 minϵ6eλ⋆(β)7 ! . (153)

    work page

  3. [3]

    (154) Proof

    The fixed point is the ground state In the β → ∞limit and the spectral gap, eλ⋆(β), of the rescaled transition matrix EγTγ · 2||HS ||(dim +1) α2t is lower bounded by a constant, giving the two limits lim β→∞ ρfix = |1⟩ ⟨1| and lim β→∞ eλ⋆(β) = 2 Z −δmint/2 0 sinc2(u)du ≥ 2.43. (154) Proof. We start by understanding the fixed points of 1+EγT (γ) on , condi...

    work page 2000

  4. [4]

    Simulated quantum computation of molecular energies

    Al´ an Aspuru-Guzik, Anthony D Dutoi, Peter J Love, and Martin Head-Gordon. Simulated quantum computation of molecular energies. Science, 309(5741):1704–1707, 2005

    work page 2005

  5. [5]

    Elucidating reaction mechanisms on quantum computers

    Markus Reiher, Nathan Wiebe, Krysta M Svore, Dave Wecker, and Matthias Troyer. Elucidating reaction mechanisms on quantum computers. Proceedings of the national academy of sciences , 114(29):7555–7560, 2017

    work page 2017

  6. [6]

    Berry, Craig Gidney, William J

    Joonho Lee, Dominic W. Berry, Craig Gidney, William J. Huggins, Jarrod R. McClean, Nathan Wiebe, and 39 Ryan Babbush. Even more efficient quantum computations of chemistry through tensor hypercontraction. PRX Quantum , 2:030305, Jul 2021

    work page 2021

  7. [7]

    Fernando G. S. L. Brand˜ ao, Amir Kalev, Tongyang Li, Cedric Yen-Yu Lin, Krysta M. Svore, and Xiaodi Wu. Quantum sdp solvers: Large speed-ups, optimality, and applications to quantum learning, 2019

    work page 2019

  8. [8]

    Sample-efficient learning of quantum many-body systems

    Anurag Anshu, Srinivasan Arunachalam, Tomotaka Kuwahara, and Mehdi Soleimanifar. Sample-efficient learning of quantum many-body systems. Nature Physics, 17(8):931–935, August 2021. arXiv:2004.07266 [cond-mat, physics:quant-ph]

    arXiv 2021

Show all 54 references
  1. [9]

    The complexity of the local hamiltonian problem, 2005

    Julia Kempe, Alexei Kitaev, and Oded Regev. The complexity of the local hamiltonian problem, 2005

    2005

  2. [10]

    Equation of state calculations by fast computing machines

    Nicholas Metropolis, Arianna W Rosenbluth, Marshall N Rosenbluth, Augusta H Teller, and Edward Teller. Equation of state calculations by fast computing machines. The journal of chemical physics , 21(6):1087–1092, 1953

    1953

  3. [11]

    On the hardness of approximate reasoning

    Dan Roth. On the hardness of approximate reasoning. Artificial Intelligence, 82(1-2):273–302, 1996

    1996

  4. [12]

    E. Y. Loh, Jr., J. E. Gubernatis, R. T. Scalettar, S. R. White, D. J. Scalapino, and R. L. Sugar. Sign problem in the numerical simulation of many-electron systems. prb, 41(13):9301–9307, May 1990

    1990

  5. [13]

    Computational complexity and fundamental limitations to fermionic quantum monte carlo simulations

    Matthias Troyer and Uwe-Jens Wiese. Computational complexity and fundamental limitations to fermionic quantum monte carlo simulations. Physical Review Letters, 94(17), May 2005

    2005

  6. [14]

    Pillai, Gareth O

    Alexandros Beskos, Natesh S. Pillai, Gareth O. Roberts, Jesus M. Sanz-Serna, and Andrew M. Stuart. Optimal tuning of the hybrid monte-carlo algorithm, 2010

    2010

  7. [15]

    A conceptual introduction to hamiltonian monte carlo, 2018

    Michael Betancourt. A conceptual introduction to hamiltonian monte carlo, 2018

    2018

  8. [16]

    Temme, T

    K. Temme, T. J. Osborne, K. G. Vollbrecht, D. Poulin, and F. Verstraete. Quantum metropolis sampling. Nature, 471(7336):87–90, March 2011

    2011

  9. [17]

    Quantum arthur–merlin games

    Chris Marriott and John Watrous. Quantum arthur–merlin games. computational complexity, 14(2):122– 152, 2005

    2005

  10. [18]

    Kastoryano, Fernando G

    Chi-Fang Chen, Michael J. Kastoryano, Fernando G. S. L. Brand˜ ao, and Andr´ as Gily´ en. Quantum thermal state preparation, 2023

    2023

  11. [19]

    Doriguello, and Michael J

    Andr´ as Gily´ en, Chi-Fang Chen, Joao F. Doriguello, and Michael J. Kastoryano. Quantum generalizations of glauber and metropolis dynamics, 2024

    2024

  12. [20]

    Ground state preparation via dynamical cooling

    Danial Motlagh, Modjtaba Shokrian Zini, Juan Miguel Arrazola, and Nathan Wiebe. Ground state preparation via dynamical cooling. arXiv preprint arXiv:2404.05810 , 2024

    2024 arXiv

  13. [21]

    Mario Motta, Chong Sun, Adrian T. K. Tan, Matthew J. O’Rourke, Erika Ye, Austin J. Minnich, Fernando G. S. L. Brand˜ ao, and Garnet Kin-Lic Chan. Determining eigenstates and thermal states on a quantum computer using quantum imaginary time evolution. Nature Physics, 16(2):205–...

    2019

  14. [22]

    Single-ancilla ground state preparation via lindbladians

    Zhiyan Ding, Chi-Fang Chen, and Lin Lin. Single-ancilla ground state preparation via lindbladians. Physical Review Research, 6(3):033147, 2024

    2024

  15. [23]

    Markovian master equations

    E Brian Davies. Markovian master equations. Communications in mathematical Physics, 39:91–110, 1974

    1974

  16. [24]

    Efficient quantum gibbs samplers with kubo–martin–schwinger detailed balance condition, 2024

    Zhiyan Ding, Bowen Li, and Lin Lin. Efficient quantum gibbs samplers with kubo–martin–schwinger detailed balance condition, 2024

    2024

  17. [25]

    Statistical-Mechanical Theory of Irreversible Processes

    Ryogo Kubo. Statistical-Mechanical Theory of Irreversible Processes. I. Journal of the Physical Society of Japan, 12(6):570–586, June 1957

    1957

  18. [26]

    Martin and Julian Schwinger

    Paul C. Martin and Julian Schwinger. Theory of Many-Particle Systems. I. Physical Review, 115(6):1342– 1373, September 1959

    1959

  19. [27]

    Dalzell, Sam McArdle, Mario Berta, Przemyslaw Bienias, Chi-Fang Chen, Andr´ as Gily´ en, Connor T

    Alexander M. Dalzell, Sam McArdle, Mario Berta, Przemyslaw Bienias, Chi-Fang Chen, Andr´ as Gily´ en, Connor T. Hann, Michael J. Kastoryano, Emil T. Khabiboulline, Aleksander Kubica, Grant Salton, Sam- son Wang, and Fernando G. S. L. Brand˜ ao. Quantum algorithms: A survey of ...

    2023

  20. [28]

    Alhambra

    Cambyse Rouz´ e, Daniel Stilck Fran¸ ca, and´Alvaro M. Alhambra. Efficient thermalization and universal quantum computing with quantum gibbs samplers, 2024

    2024

  21. [29]

    Preparing thermal states on noiseless and noisy programmable quantum processors, 2023

    Oles Shtanko and Ramis Movassagh. Preparing thermal states on noiseless and noisy programmable quantum processors, 2023

    2023

  22. [30]

    Probabilistic inference using markov chain monte carlo methods

    Radford M Neal. Probabilistic inference using markov chain monte carlo methods. 1993

    1993

  23. [31]

    Hoffman and Andrew Gelman

    Matthew D. Hoffman and Andrew Gelman. The no-u-turn sampler: Adaptively setting path lengths in hamiltonian monte carlo, 2011

    2011

  24. [32]

    Quantum collision models: Open system dynamics from repeated interactions

    Francesco Ciccarello, Salvatore Lorenzo, Vittorio Giovannetti, and G Massimo Palma. Quantum collision models: Open system dynamics from repeated interactions. Physics Reports, 954:1–70, 2022

    2022

  25. [33]

    Quantum digital cooling

    Stefano Polla, Yaroslav Herasymenko, and Thomas E O’Brien. Quantum digital cooling. Physical Review A, 104(1):012414, 2021

    2021

  26. [34]

    Equilibrium and nonequilibrium steady states with the repeated interaction protocol: Relaxation dynamics and energetic cost

    Alessandro Prositto, Madeline Forbes, and Dvira Segal. Equilibrium and nonequilibrium steady states with the repeated interaction protocol: Relaxation dynamics and energetic cost. arXiv preprint arXiv:2501.05392, 2025. 40

    2025 arXiv

  27. [35]

    Quantum simulation of lindbladian dynamics via repeated interactions, 2024

    Matthew Pocrnic, Dvira Segal, and Nathan Wiebe. Quantum simulation of lindbladian dynamics via repeated interactions, 2024

    2024

  28. [36]

    Markov chains and mixing times , volume 107

    David A Levin and Yuval Peres. Markov chains and mixing times , volume 107. American Mathematical Soc., 2017

    2017

  29. [37]

    General mixing time bounds for finite markov chains via the absolute spectral gap

    Daniel Jerison. General mixing time bounds for finite markov chains via the absolute spectral gap. arXiv preprint arXiv:1310.8021, 2013

    2013 arXiv

  30. [38]

    Supercooling and the mpemba effect: When hot water freezes quicker than cold

    David Auerbach. Supercooling and the mpemba effect: When hot water freezes quicker than cold. Amer- ican Journal of Physics , 63(10):882–885, 1995

    1995

  31. [39]

    Cool? Physics Education, 4(3):172, May 1969

    E B Mpemba and D G Osborne. Cool? Physics Education, 4(3):172, May 1969

    1969

  32. [40]

    Hyperacceleration of quantum thermalization dynamics by bypassing long-lived coherences: An analytical treatment

    Felix Ivander, Nicholas Anto-Sztrikacs, and Dvira Segal. Hyperacceleration of quantum thermalization dynamics by bypassing long-lived coherences: An analytical treatment. Phys. Rev. E , 108:014130, Jul 2023

    2023

  33. [41]

    Nonequilibrium thermodynamics of the markovian mpemba effect and its inverse

    Zhiyue Lu and Oren Raz. Nonequilibrium thermodynamics of the markovian mpemba effect and its inverse. Proceedings of the National Academy of Sciences , 114(20):5083–5088, 2017

    2017

  34. [42]

    Observation of quantum strong mpemba effect

    Jie Zhang, Gang Xia, Chun-Wang Wu, Ting Chen, Qian Zhang, Yi Xie, Wen-Bo Su, Wei Wu, Cheng-Wei Qiu, Ping-Xing Chen, et al. Observation of quantum strong mpemba effect. Nature Communications, 16(1):301, 2025

    2025

  35. [43]

    Openfermion: the electronic structure package for quantum computers

    Jarrod R McClean, Nicholas C Rubin, Kevin J Sung, Ian D Kivlichan, Xavier Bonet-Monroig, Yudong Cao, Chengyu Dai, E Schuyler Fried, Craig Gidney, Brendan Gimby, et al. Openfermion: the electronic structure package for quantum computers. Quantum Science and Technology, 5(3):034...

    2020

  36. [44]

    Berkelbach, Nick S

    Qiming Sun, Timothy C. Berkelbach, Nick S. Blunt, George H. Booth, Sheng Guo, Zhendong Li, Junzi Liu, James D. McClain, Elvira R. Sayfutyarova, Sandeep Sharma, Sebastian Wouters, and Garnet Kin-Lic Chan. Pyscf: the python-based simulations of chemistry framework. WIREs Computa...

    2018

  37. [45]

    Preparing random states and benchmarking with many-body quantum chaos

    Joonhee Choi, Adam L Shaw, Ivaylo S Madjarov, Xin Xie, Ran Finkelstein, Jacob P Covey, Jordan S Cotler, Daniel K Mark, Hsin-Yuan Huang, Anant Kale, et al. Preparing random states and benchmarking with many-body quantum chaos. Nature, 613(7944):468–473, 2023

    2023

  38. [46]

    Hamiltonian simulation by qubitization

    Guang Hao Low and Isaac L Chuang. Hamiltonian simulation by qubitization. Quantum, 3:163, 2019

    2019

  39. [47]

    Theory of trotter error with commutator scaling

    Andrew M Childs, Yuan Su, Minh C Tran, Nathan Wiebe, and Shuchen Zhu. Theory of trotter error with commutator scaling. Physical Review X , 11(1):011020, 2021

    2021

  40. [48]

    The clifford group forms a unitary 3-design

    Zak Webb. The clifford group forms a unitary 3-design. arXiv preprint arXiv:1510.02769 , 2015

    2015 arXiv

  41. [49]

    Composite quantum simulations

    Matthew Hagan and Nathan Wiebe. Composite quantum simulations. Quantum, 7:1181, 2023

    2023

  42. [50]

    Brand˜ ao, Wissam Chemissany, Nicholas Hunter-Jones, Richard Kueng, and John Preskill

    Fernando G.S.L. Brand˜ ao, Wissam Chemissany, Nicholas Hunter-Jones, Richard Kueng, and John Preskill. Models of quantum complexity growth. PRX Quantum , 2:030316, Jul 2021

    2021

  43. [51]

    Relations for certain symmetric norms and anti-norms before and after partial trace

    Alexey E Rastegin. Relations for certain symmetric norms and anti-norms before and after partial trace. Journal of Statistical Physics , 148:1040–1053, 2012. Appendix A: T echnical Proofs

    2012

  44. [52]

    For sinc2 xt 2 and δmin as defined in Eq

    Sinc Bounds Lemma 14 (Sinc Function Bounds) . For sinc2 xt 2 and δmin as defined in Eq. (3), we will make significant use of the following bounds: |x| ≥δmin =⇒ sinc2 xt 2 ≤ 4 δ2 mint2 (A1) |x| ≤ √ 2 t =⇒ sinc2 xt 2 ≥ 1 − |x|2t2 2 . (A2) Proof. The first inequality is rather tr...

  45. [53]

    Lemmas 16 and 17 are used to compute the effects of the randomized interactions in a form that are usable in the main result of Lemma 2

    Haar Integral Proofs In this section we present the more technical work needed to state our results in Section II. Lemmas 16 and 17 are used to compute the effects of the randomized interactions in a form that are usable in the main result of Lemma 2. Lemma 15 can be derived f...

  46. [54]

    First we note that although RΦ(ρ) = α3 6 ∂3 αΦ(ρ) α=α⋆ for a specific α⋆ > 0 our proof will hold for any α⋆

    W eak-Coupling Remainder Bound Proof of Theorem 9. First we note that although RΦ(ρ) = α3 6 ∂3 αΦ(ρ) α=α⋆ for a specific α⋆ > 0 our proof will hold for any α⋆. To compute the trace norm we will use the triangle inequality, unitary invariance of the Sch¨ atten norms, and submul...

Pith tools

Reviewed August 9, 2026 · model on record in the stance chip above.