Pith. sign in

REVIEW 2 major objections 5 minor 31 references

Chemical Potential and Analytic Continuation for Non-Hermitian Lattice Fermions

T0 review · 2 major / 5 minor · reviewed 2026-08-14 · deepseek-v4-flash

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

Pith's one-line read For a solvable 1D non-Hermitian lattice fermion, analytic continuation from imaginary to real chemical potential is valid on a finite lattice but fails after the continuum limit.

desk verdict Useful exact benchmark for non-Hermitian lattice fermions, but the AI-continuation section is a known-ansatz least-squares fit, not a general method. read the letter →

arxiv 2608.08862 v1 pith:7BISSL2Z submitted 2026-08-09 hep-lat

Chen-Te Ma , Hui Zhang This is my paper · ORCID
classification hep-lat
keywords non-HermitianlatticefermionschemicalpotentialanalyticcontinuationsignproblemHybridMonteCarloexactpropagatorimaginaryphysics-constrainedneuralnetwork
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 studies a one-dimensional free Dirac fermion on a lattice, discretized with non-Hermitian one-sided derivatives and a chemical potential, because every quantity can be computed exactly. It derives the exact propagator in both the naive and exponential chemical-potential implementations and shows that at finite lattice spacing the correlator is analytic in the chemical potential, while the infinite-volume or continuum limit can destroy that analyticity. It then extends Hybrid Monte Carlo to finite chemical potential by pairing degenerate flavors with chemical potentials $(\mu,-\mu)$ or $(i\mu,i\mu)$, for which the fermion determinant is the absolute square of a single determinant and the sign problem disappears. Finally, it uses a physics-constrained neural network and a Laurent exponential ansatz to continue two-point correlators from imaginary to real chemical potential, reproducing the exact results from few training points.

What carries the argument

The load-bearing object is the paired forward/backward finite-difference Dirac operator on a finite one-dimensional lattice. The forward operator $D_+(\mu_1)$ uses a forward shift $e^{-a\mu_1}\gamma^1$, and the backward operator $D_-(\mu_2)$ uses a backward shift with $e^{a\mu_2}\gamma^1$, chosen so that $D_-(\mu_2)=-D_+(\mu_1)^\dagger$ for the two pairings $(\mu,-\mu)$ and $(i\mu,i\mu)$. This conjugacy relation collapses the two-flavor determinant to an absolute square and makes the pseudofermion action Hermitian. The exact propagator is obtained by contour integration over the lattice momentum poles, giving closed-form rational expressions whose $\mu$-dependence is analytic for finite $N_t$; the continuum limit then selects different decay branches, producing the non-analyticity.

What would settle it

Compute the exact finite-lattice propagator at fixed lattice spacing for a chemical potential with $\operatorname{Re}\mu>m$, and take the $N_t\to\infty$ limit at fixed $n=x/a$; if the limiting correlator is still an analytic function of $\mu$ rather than switching branches at $\operatorname{Re}(m-\mu)=0$, the paper's main non-commutation claim fails.

Watch

Extended reading notes

Core claim

The central discovery is that neither the sign problem nor analytic-continuation failure is intrinsic to non-Hermitian lattice fermions with a chemical potential. For the forward-difference Dirac operator $D_+(\mu)$, the exact lattice propagator has the same functional form for any value of $\mu$ while the lattice size $N_t$ is finite; analyticity in $\mu$ is lost only when $N_t\to\infty$ at fixed lattice spacing, with the correlator switching between exponential decay branches according to whether $\operatorname{Re}(m\mp\mu)$ is positive or negative. In the exponential prescription the chemical potential decouples from the continuum limit, which requires only $ma\to0$. For two degenerate flavors with $D_-(\mu_2)=-D_+(\mu_1)^\dagger$, the determinant identity $\det(D_++m)\det(-D_+^\dagger+m)=|\det(D_++m)|^2$ holds for the pairings $(\mu,-\mu)$ and $(i\mu,i\mu)$, making the partition function non-negative and Hybrid Monte Carlo applicable. The measured two-point pseudofermion correlators agree with the exact solution, and the AI-assisted continuation from imaginary to real $\mu$ matches the exact result.

Load-bearing premise

The AI-assisted analytic-continuation benchmark presumes that the correlator lies in the three-mode space spanned by $e^{-z}$, $1$, and $e^z$, an assumption known to hold only because the exact solution was derived beforehand.

Editorial extensions

If this is right

  • On a finite lattice, observables such as the two-point correlator are analytic in the chemical potential, so imaginary-to-real continuation is legitimate before the limits are taken.
  • The continuum limit and analytic continuation do not generally commute; simulations that use imaginary chemical potentials should state the order of limits explicitly.
  • The exponential chemical-potential prescription needs only $ma\to0$ to recover the continuum theory, so it is a safer lattice definition than the naive additive insertion of $\mu$.
  • For two degenerate flavors with $(\mu_1,\mu_2)=(\mu,-\mu)$ or $(i\mu,i\mu)$, Hybrid Monte Carlo is sign-problem-free and reproduces the exact two-point pseudofermion correlators at $N_t=16$ and $32$.
  • The sign problem in these configurations coincides with Dirac eigenvalues of negative real part and reflects a technical obstruction to pseudofermion sampling, not a physical breakdown of the fermionic theory.

Reading between the lines

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

  • If the eigenvalue criterion persists in interacting theories, monitoring the minimal real part of the Dirac spectrum during a simulation could serve as an inexpensive early warning for sign-problem onset.
  • The absolute-square determinant identity suggests a constructive recipe for other pairings: choose flavor chemical potentials so that $D_-(\mu_2)=-D_+(\mu_1)^\dagger$; testing twisted-mass-like pairings for non-degenerate masses is an immediate next step the paper leaves open.
  • The success of the three-mode ansatz is partly circular, since the ansatz was chosen because the exact solution was already known; a practical method would need an independent criterion to select the mode set from the lattice operator's pole structure.
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 / 5 minor

Summary. This paper studies one-dimensional free lattice Dirac fermions with a non-Hermitian one-sided derivative and a chemical potential, in both a naive and an exponential implementation. It derives exact finite-lattice propagators and shows that they are analytic in the chemical potential away from poles, while the infinite-volume/continuum limit can destroy analyticity, so the limits do not commute. For two degenerate flavors with paired chemical potentials (mu,-mu) or (i mu,i mu), the pseudofermion determinant is shown to be non-negative, and Hybrid Monte Carlo simulations reproduce the exact correlators for N_t=16,32. The final section presents an 'AI-assisted' analytic continuation from imaginary to real chemical potential using a three-mode Laurent exponential model, a physics-constrained neural network, and a hybrid residual network.

Significance. If the central claims hold, the exact-solution part provides a clean solvable demonstration that finite-volume lattice observables can be analytic in the chemical potential while the continuum limit is not, clarifying an important subtlety for imaginary-chemical-potential methods. The determinant identity (28) extends earlier work on non-Hermitian lattice fermions and yields a sign-problem-free HMC formulation for the paired chemical-potential assignments, with numerical agreement on two lattice sizes. The AI-assisted continuation is the weakest part: as the paper itself acknowledges, the three-parameter ansatz is chosen using the exact solution, the hybrid residual network is never activated, and the method reduces to a linear least-squares fit of a known functional form. The paper is honest about these limitations in Sec. 6, but the abstract and Sec. 1.1 overstate the benchmark value.

major comments (2)
  1. [Sec. 6.1, Eqs. (32)-(37)] The three-mode hypothesis space H = span{e^{-z},1,e^z} is fixed by the exact solution derived in Sec. 2; the imaginary-axis data for O_{j,j+1}^{1,1} and O_{j,j+1}^{2,2} are single exponentials C exp(-i mu), so the ansatz contains the answer by construction. No model-selection criterion is provided for the truncation order K in Eq. (43) when the exact form is unavailable. The agreement in Figs. 10-13 therefore demonstrates only that a three-parameter least-squares fit reproduces a known functional form; it does not validate an unsupervised AI-assisted analytic-continuation method. This undercuts the benchmark claim in Sec. 1.1 ('providing a benchmark for the method across different lattice sizes') and in Sec. 6.3 ('demonstrate the usefulness of the AI-assisted fitting framework'). The authors should either supply a selection procedure that does not use the exact solution (e.g., cross-validated choice of K with leave-out tests) or explicitly reframe Sec. 6 as an illustration conditional on the known solution.
  2. [Sec. 6.2.4, Eqs. (53)-(54)] The adaptive gate uses epsilon = 10^{-10}, while the PCNN achieves a normalized RMSE of approximately 10^{-13}, so the residual network f_hyb - f_phys is never activated in any benchmark. The paper's own text states that the study 'validates the design of the hybrid framework rather than demonstrating the practical improvement,' but the abstract and the bullet in Sec. 1.1 present AI-assisted continuation as a demonstrated result. To support the claimed benchmark, the authors should either test the hybrid model on data where the physics backbone is incomplete or noisy (as they suggest in Sec. 6.2.4) or revise the abstract and summary to state that the hybrid component is untested.
minor comments (5)
  1. [Sec. 3, Eq. (28)] The identity uses a matrix gamma5 without defining it; in this 1D (or 1+1D) setting the authors should specify the explicit matrix (e.g., sigma_x or gamma^3) and note that it anticommutes with gamma1.
  2. [Sec. 4, Figs. 4 and 5] The caption 'When the simulation suffers from the sign problem...' is attached to panels that include sign-problem-free assignments; for example, Fig. 4 lower-left is (mu1,mu2)=(mu,-mu), which is sign-problem-free, and its minimum real part is positive. The captions should distinguish the sign-problem and sign-problem-free panels.
  3. [Sec. 2.3, Eq. (18)] The expression '=e nµa' should be written with proper exponents, e.g., e^{n mu a}, and the convention for n for x<0 should be stated explicitly.
  4. [Sec. 6.1, Eq. (33)] The notation fPCNN(z) = phi(z) w is ambiguous because phi(z) is given as a row vector; define the inner-product convention so that the three trainable weights are unambiguously associated with the modes.
  5. [Sec. 1.1 and Sec. 6.1] The term 'AI-assisted fitting' overstates the method presented: the PCNN is equivalent to a three-parameter linear least-squares fit, as the paper itself states in Sec. 6.1. A more neutral term such as 'physics-constrained fit' would better match the content.

Circularity Check

1 steps flagged · score 4.0 of 10

AI-continuation benchmark is circular: the three-mode ansatz is taken from the exact solution, so the real-axis 'prediction' is the fitted function evaluated on the real axis.

  1. fitted input called prediction [Sec. 6.1, Eqs. (32)-(37); Sec. 6.3]
    "To exploit the analytic structure of the lattice solution, we construct a physics-constrained neural network (PCNN) whose hidden features are prescribed by the analytic approximation rather than learned from data. ... the analytic feature layer is chosen as ϕ(z) = [e−z 1e z] ... shows that both formulations span the same hypothesis space, H = span{e−z, 1, ez}. ... After training, we perform analytic continuation by replacing exp(−iµ) with exp(−µ)."

    The real-axis result is obtained by fitting the coefficients of H = span{e^{-z}, 1, e^z} to imaginary-axis data and then evaluating the fitted function at real μ. Since the hypothesis space was chosen from the exact lattice solution ('the three dominant exponential modes predicted by the lattice analysis'), the ansatz already contains the target by construction. The paper gives no data-driven criterion for selecting K or the basis, and its own limitation note concedes 'dependence on prior physical knowledge.' Thus the benchmark validates a three-parameter fit on a known functional form rather than an unsupervised continuation method. The exact-solution and HMC results are unaffected.

full rationale

The exact derivation in Sec. 2 (Eqs. 11-21) and the determinant identity in Sec. 3 (Eq. 28) are self-contained: the finite-lattice propagator is computed directly, the infinite-volume limits are taken consistently, and the HMC sign-problem-free condition follows from the identity D−(μ2) = −[D+(μ1)]^† and γ5 conjugation, not from a fitted parameter or a self-citation chain. The self-citations [18,19,25] are contextual and not load-bearing because the needed determinant identity is proved in-text. The circularity concern is confined to Sec. 6: the PCNN/Laurent model fixes the hypothesis space H = span{e^{-z}, 1, e^z} using the exact solution as prior knowledge, fits three coefficients to imaginary-axis data, and reports the fitted function evaluated at real μ as the continuation. The exact real-axis values are not used in the fit, so this is a legitimate consistency check; but because the ansatz was selected from the exact solution, the demonstration does not establish that the method can select a correct functional form when the answer is unknown. The paper's own limitation statements (Sec. 6.1 and Sec. 6.2.4) acknowledge this. This warrants a moderate circularity score, not a high one.

Assumptions & free parameters 2 free parameters · 4 assumptions · 0 invented entities

The exact-solution derivations use standard contour integration and geometric series; the HMC sign-problem-free result is a linear-algebra identity for the paired potentials; the AI fitting introduces fitted coefficients and an ansatz informed by the exact solution, which is the main source of circularity burden.

free parameters (2)
  • AI fit coefficients a_{-1}, a_0, a_1 = e.g., -0.4375 for O^{1,1}_{j,j+1} at N_t=16; -1.3125 for O^{2,2}_{j,j+1} at N_t=16
    In Sec. 6.3 the Laurent model and PCNN coefficients are fitted to imaginary-chemical-potential data by least squares; the real-axis continuation is this fitted function evaluated at real μ.
  • Regularization coefficient λ and gate threshold ε = λ tunable; ε=10^{-10}
    Chosen by hand in Sec. 6.2.2 and 6.2.3; they do not affect the reported results because the adaptive gate is inactive.
assumptions (4)
  • domain assumption The forward and backward one-sided derivative lattice action with anti-periodic boundary conditions is a valid regularization of the 1D Dirac fermion.
    Used throughout Sec. 2 for the exact solution; the continuum limit is taken as a→0 with ma and μa small.
  • standard math The mode sums are evaluated using the residue theorem and geometric-series identities.
    Used in Sec. 2.2 and 2.3 to derive the exact propagators, Eqs. (11)-(12) and (18)-(19).
  • domain assumption The exponential (Hasenfratz-Karsch) chemical-potential prescription reduces to the continuum action when ma→0.
    Adopted in Sec. 2.3 with citation [27]; the conclusion that the exponential form needs no additional μ constraint relies on this.
  • domain assumption The correlator is analytic in μ on the finite lattice and is representable by the three-mode exponential ansatz.
    The AI continuation in Sec. 6 selects the ansatz {e^{-z},1,e^z} because the exact solution has that form; no independent criterion is given.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Chemical Potential and Analytic Continuation for Non-Hermitian Lattice Fermions." pith.science (2026). https://pith.science/paper/7BISSL2Z

@misc{pith2026260808862,
  author       = {Pith},
  title        = {Pith review of: Chemical Potential and Analytic Continuation for Non-Hermitian Lattice Fermions},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/7BISSL2Z}},
  note         = {Machine review of arXiv:2608.08862}
}
abstract

We introduce a chemical potential for non-Hermitian lattice fermions and show that, for even flavors with degenerate masses and paired chemical potentials $(\mu,-\mu)$ or $(i\mu,i\mu)$, the Hybrid Monte Carlo algorithm is free of the sign problem. For one-dimensional free fermions, we demonstrate that the sign problem is a numerical rather than physical obstruction and derive the exact propagator, which is analytic at finite lattice spacing away from its poles but becomes non-analytic in the continuum limit. Finally, we use AI-assisted fitting to perform analytic continuation from imaginary to real chemical potentials.

Figures

Figures reproduced from arXiv: 2608.08862 by the authors.

Figure 1
Figure 1. Closed contour C1 in the complex w-plane. The contour is oriented counterclockwise and encloses the Nt simple poles located at w = (2j + 1)/2 with j = 0, 1, · · · , Nt − 1, namely w = 1/2, 3/2, · · · , Nt − 1/2. 2.3 Exponential Form We now consider the lattice theory obtained from the forward finite-difference SF = a N Xt−1 n=0 ψ¯(n)  γ1 e −aµ1ψ(n + 1) − ψ(n) a + mψ(n)  = a X n1,n2;α1,α2 ψ¯(n1)α1 [PITH_FULL_IMAGE… view at source ↗
Figure 2
Figure 2. The smallest real part of eigenvalues of a Dirac matrix corresponding to the [PITH_FULL_IMAGE:figures/full_fig_p014_2.png] view at source ↗
Figure 3
Figure 3. The smallest real part of eigenvalues of a Dirac matrix corresponding to the [PITH_FULL_IMAGE:figures/full_fig_p015_3.png] view at source ↗
Figures from the paper (10 more)
Figure 4
Figure 4. Figure 4: When the simulation suffers from the sign problem, the smallest real part of [PITH_FULL_IMAGE:figures/full_fig_p016_4.png]
Figure 5
Figure 5. Figure 5: When the simulation suffers from the sign problem, the smallest real part [PITH_FULL_IMAGE:figures/full_fig_p017_5.png]
Figure 6
Figure 6. Figure 6: We use the Hybrid Monte Carlo (HMC) algorithm to compute [PITH_FULL_IMAGE:figures/full_fig_p018_6.png]
Figure 7
Figure 7. Figure 7: We use the Hybrid Monte Carlo (HMC) algorithm to compute Re( [PITH_FULL_IMAGE:figures/full_fig_p019_7.png]
Figure 8
Figure 8. Figure 8: We use the Hybrid Monte Carlo (HMC) algorithm to compute Im( [PITH_FULL_IMAGE:figures/full_fig_p020_8.png]
Figure 9
Figure 9. Figure 9: Architecture of the proposed physics-constrained neural network (PCNN). [PITH_FULL_IMAGE:figures/full_fig_p022_9.png]
Figure 10
Figure 10. Figure 10: Analytic continuation from imaginary chemical potential (left panels) to real [PITH_FULL_IMAGE:figures/full_fig_p031_10.png]
Figure 11
Figure 11. Figure 11: Analytic continuation from imaginary chemical potential (left panels) to real [PITH_FULL_IMAGE:figures/full_fig_p032_11.png]
Figure 12
Figure 12. Figure 12: Analytic continuation from imaginary chemical potential (left panels) to real [PITH_FULL_IMAGE:figures/full_fig_p033_12.png]
Figure 13
Figure 13. Figure 13: Analytic continuation from imaginary chemical potential (left panels) to real [PITH_FULL_IMAGE:figures/full_fig_p034_13.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

31 extracted references · 8 canonical work pages

  1. [18]

    Naive lattice fermion without dou- blers,

    X. Guo, C. T. Ma and H. Zhang, “Naive lattice fermion without dou- blers,” Phys. Rev. D104, no.9, 094505 (2021) doi:10.1103/PhysRevD.104.094505 [arXiv:2105.10977 [hep-lat]]

    arXiv 2021

  2. [1]

    Space-time approach to nonrelativistic quantum mechanics,

    R. P. Feynman, “Space-time approach to nonrelativistic quantum mechanics,” Rev. Mod. Phys.20, 367-387 (1948) doi:10.1103/RevModPhys.20.367

    doi:10.1103/revmodphys.20.367 1948

  3. [2]

    Confinement of Quarks,

    K. G. Wilson, “Confinement of Quarks,” Phys. Rev. D10, 2445-2459 (1974) doi:10.1103/PhysRevD.10.2445

    doi:10.1103/physrevd.10.2445 1974

  4. [3]

    Methods of contemporary gauge theory,

    Y. Makeenko, “Methods of contemporary gauge theory,” Cambridge Univer- sity Press, 2005, ISBN 978-0-521-02215-6, 978-0-521-80911-5, 978-0-511-05768-7 doi:10.1017/CBO9780511535147

    work page doi:10.1017/cbo9780511535147 2005

  5. [4]

    Quantum chromodynamics on the lattice,

    C. Gattringer and C. B. Lang, “Quantum chromodynamics on the lattice,” Lect. Notes Phys.788, 1-343 (2010) Springer, 2010, ISBN 978-3-642-01849-7, 978-3-642- 01850-3 doi:10.1007/978-3-642-01850-3 35

    doi:10.1007/978-3-642-01850-3 2010

  6. [5]

    Lattice chiral fermion without Hermiticity,

    C. T. Ma and H. Zhang, “Lattice chiral fermion without Hermiticity,” Int. J. Mod. Phys. A40, no.23, 2530008 (2025) doi:10.1142/S0217751X2530008X [arXiv:2411.09886 [hep-lat]]

    arXiv 2025

  7. [6]

    Absence of Neutrinos on a Lattice. 1. Proof by Homotopy Theory,

    H. B. Nielsen and M. Ninomiya, “Absence of Neutrinos on a Lattice. 1. Proof by Homotopy Theory,” Nucl. Phys. B185, 20 (1981) [erratum: Nucl. Phys. B195, 541 (1982)] doi:10.1016/0550-3213(82)90011-6

    doi:10.1016/0550-3213(82)90011-6 1981

  8. [7]

    Absence of Neutrinos on a Lattice. 2. Intu- itive Topological Proof,

    H. B. Nielsen and M. Ninomiya, “Absence of Neutrinos on a Lattice. 2. Intu- itive Topological Proof,” Nucl. Phys. B193, 173-194 (1981) doi:10.1016/0550- 3213(81)90524-1

    doi:10.1016/0550- 1981

Show all 31 references
  1. [8]

    Lattice Fermions in Euclidean Space-time,

    L. H. Karsten, “Lattice Fermions in Euclidean Space-time,” Phys. Lett. B104, 315-319 (1981) doi:10.1016/0370-2693(81)90133-7

    1981 doi

  2. [9]

    Properties of Bethe-Salpeter Wave Functions,

    G. C. Wick, “Properties of Bethe-Salpeter Wave Functions,” Phys. Rev.96, 1124- 1134 (1954) doi:10.1103/PhysRev.96.1124

    1954 doi

  3. [10]

    Can Euclidean lattice quantum field theory be analytically continued into Minkowski space?,

    B. P. Kosyakov, E. Y. Popov and M. A. Vronski˘i, “Can Euclidean lattice quantum field theory be analytically continued into Minkowski space?,” Mod. Phys. Lett. A 41, no.26, 2650146 (2026) doi:10.1142/s0217732326501464 [arXiv:2605.18787 [hep- th]]

    2026 arXiv

  4. [11]

    Hamiltonian Formulation of Wilson’s Lattice Gauge Theories,

    J. B. Kogut and L. Susskind, “Hamiltonian Formulation of Wilson’s Lattice Gauge Theories,” Phys. Rev. D11, 395-408 (1975) doi:10.1103/PhysRevD.11.395

    1975 doi

  5. [12]

    A Remnant of Chiral Symmetry on the Lattice,

    P. H. Ginsparg and K. G. Wilson, “A Remnant of Chiral Symmetry on the Lattice,” Phys. Rev. D25, 2649 (1982) doi:10.1103/PhysRevD.25.2649

    1982 doi

  6. [13]

    Exact chiral symmetry on the lattice and the Ginsparg-Wilson re- lation,

    M. Luscher, “Exact chiral symmetry on the lattice and the Ginsparg-Wilson re- lation,” Phys. Lett. B428, 342-345 (1998) doi:10.1016/S0370-2693(98)00423-7 [arXiv:hep-lat/9802011 [hep-lat]]. 36

    1998 arXiv

  7. [14]

    More about exactly massless quarks on the lattice,

    H. Neuberger, “More about exactly massless quarks on the lattice,” Phys. Lett. B427, 353-355 (1998) doi:10.1016/S0370-2693(98)00355-4 [arXiv:hep-lat/9801031 [hep-lat]]

    1998 arXiv

  8. [15]

    A New formulation of lattice gauge theory with fermions,

    I. O. Stamatescu and T. T. Wu, “A New formulation of lattice gauge theory with fermions,” CERN-TH-6631-92

  9. [16]

    Lattice fermion formulation with one-sided derivatives,

    I. O. Stamatescu and T. T. Wu, “Lattice fermion formulation with one-sided derivatives,” Nucl. Phys. B Proc. Suppl.42, 838-840 (1995) doi:10.1016/0920- 5632(95)00397-R

    1995 doi

  10. [17]

    Continuum behavior of lattice QED, discretized with one sided lattice differences, in one loop order,

    N. Sadooghi and H. J. Rothe, “Continuum behavior of lattice QED, discretized with one sided lattice differences, in one loop order,” Phys. Rev. D55, 6749-6759 (1997) doi:10.1103/PhysRevD.55.6749 [arXiv:hep-lat/9610001 [hep-lat]]

    1997 arXiv

  11. [19]

    Non-Hermitian lattice fermions in the 2D Gross-Neveu-Yukawa model,

    X. Guo, C. T. Ma and H. Zhang, “Non-Hermitian lattice fermions in the 2D Gross-Neveu-Yukawa model,” Phys. Rev. D110, no.3, 034502 (2024) doi:10.1103/PhysRevD.110.034502 [arXiv:2404.18441 [hep-th]]

    2024 arXiv

  12. [20]

    ON THE EUCLIDEAN STRUCTURE OF RELATIVIS- TIC FIELD THEORY,

    J. Schwinger, “ON THE EUCLIDEAN STRUCTURE OF RELATIVIS- TIC FIELD THEORY,” Proc. Nat. Acad. Sci.44, no.9, 956-965 (1958) doi:10.1073/pnas.44.9.956

    1958 doi

  13. [21]

    Axioms for Euclidean Green’s Functions. 2.,

    K. Osterwalder and R. Schrader, “Axioms for Euclidean Green’s Functions. 2.,” Commun. Math. Phys.42, 281 (1975) doi:10.1007/BF01608978

    1975 doi

  14. [22]

    On the Green’s functions of quantized fields. 1.,

    J. S. Schwinger, “On the Green’s functions of quantized fields. 1.,” Proc. Nat. Acad. Sci.37, 452-455 (1951) doi:10.1073/pnas.37.7.452 37

    1951 doi

  15. [23]

    Finite density (might well be easier) at finite temperature,

    M. P. Lombardo, “Finite density (might well be easier) at finite temperature,” Nucl. Phys. B Proc. Suppl.83, 375-377 (2000) doi:10.1016/S0920-5632(00)91678-5 [arXiv:hep-lat/9908006 [hep-lat]]

    2000 arXiv

  16. [24]

    Finite density QCD via imaginary chemical potential,

    M. D’Elia and M. P. Lombardo, “Finite density QCD via imaginary chemical potential,” Phys. Rev. D67, 014505 (2003) doi:10.1103/PhysRevD.67.014505 [arXiv:hep-lat/0209146 [hep-lat]]

    2003 arXiv

  17. [25]

    Imaginary polarization as a way to surmount the sign problem inAb Initiocal- culations of spin-imbalanced Fermi gases,

    J. Braun, J. W. Chen, J. Deng, J. E. Drut, B. Friman, C. T. Ma and Y. D. Tsai, “Imaginary polarization as a way to surmount the sign problem inAb Initiocal- culations of spin-imbalanced Fermi gases,” Phys. Rev. Lett.110, 130404 (2013) doi:10.1103/PhysRevLett.110.130404 [arXiv:...

    2013 arXiv

  18. [26]

    QCD phase diagram for nonzero isospin-asymmetry,

    B. B. Brandt, G. Endrodi and S. Schmalzbauer, “QCD phase diagram for nonzero isospin-asymmetry,” Phys. Rev. D97, no.5, 054514 (2018) doi:10.1103/PhysRevD.97.054514 [arXiv:1712.08190 [hep-lat]]

    2018 arXiv

  19. [27]

    Chemical Potential on the Lattice,

    P. Hasenfratz and F. Karsch, “Chemical Potential on the Lattice,” Phys. Lett. B 125, 308-310 (1983) doi:10.1016/0370-2693(83)91290-X

    1983 doi

  20. [28]

    Target space duality in string theory,

    A. Giveon, M. Porrati and E. Rabinovici, “Target space duality in string theory,” Phys. Rept.244, 77-202 (1994) doi:10.1016/0370-1573(94)90070-1 [arXiv:hep- th/9401139 [hep-th]]

    1994

  21. [29]

    Parity Anomaly and Duality Web,

    C. T. Ma, “Parity Anomaly and Duality Web,” Fortsch. Phys.66, no.8-9, 1800045 (2018) doi:10.1002/prop.201800045 [arXiv:1802.08959 [hep-th]]

    2018 arXiv

  22. [30]

    AdS 3 Einstein gravity and boundary description: pedagogical re- view,

    C. T. Ma, “AdS 3 Einstein gravity and boundary description: pedagogical re- view,” Class. Quant. Grav.41, no.2, 023001 (2024) doi:10.1088/1361-6382/ad17f0 [arXiv:2310.04665 [hep-th]]

    2024 arXiv

  23. [31]

    Non-Commutative Geometry for D-Branes in Large R-R Field Back- ground,

    C. T. Ma, “Non-Commutative Geometry for D-Branes in Large R-R Field Back- ground,” [arXiv:2510.24998 [hep-th]]. 38

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.