Pith. Sign in

REVIEW

Lattice QCD at Finite Density -- An introductory review

PaperPDF from arXiv

data files

ReviewThe objections and verdict, as JSON Paper breakdownThe authors' claims, conditions and measurements, as JSON Signed receiptProof this review has not been altered
X Bluesky LinkedIn Reddit Hacker News
Shin Muroya, Atsushi Nakamura, Chiho Nonaka, Tetsuya Takaishi

High Energy Physics - Lattice · Submitted 24 Jun 2003 · 117 references

Pith's verdict: A solid, honest 2003 map of finite-density lattice QCD; its soft spots are compression in the abstract and one figure resting on an unpublished source, not the physics.

T0 · AI review

How this review was made: An AI referee (selfhosted/deepseek-v4.1-flash) read the paper on November 11, 2026 and answered a fixed list of questions about it. T0 is the starting level every paper gets. The levels describe how deeply Pith has checked the paper, never the authors.

  1. T0 AI review An AI read the paper and answered a fixed list of questions.
  2. T1 Audit Automatic checks that the references and files the paper relies on exist and match.
  3. T2 Compute The paper's code or calculations re-run with the same results.
  4. T3 Formal statement The claims written out in a formal math language a computer can read.
  5. T4 Proof Those claims proved, and the proof checked by computer.

How Pith checks papers →

· 3 major objections · 9 minor · Ask about this paper · The compiled record · Read the full review →
Challenge this review Comment (0) Read on arXiv ↗
Cite ▾
arXiv IDhep-lat/0306031 (version 1)
Pith IDOSTBCBVJ, Pith's permanent record of this paper and its review
CitationAtsushi Nakamura, Chiho Nonaka, Shin Muroya, Tetsuya Takaishi. "Lattice QCD at Finite Density -- An introductory review." arXiv:hep-lat/0306031 (2003). https://pith.science/paper/OSTBCBVJ
Cite this reviewBibTeX for Pith's review of this paper
Claim paper

Why Pith says this

Highlight any text to comment on it.

Referee report

How Pith checks papers → · Full audit of this paper →

The paper

The paper claims

A review argues that lattice QCD at finite density is organized by a single obstacle — the complex fermion determinant — and maps the ways around it.

The reading

This review sets out to give a newcomer the minimum needed to read the lattice literature on QCD at nonzero baryon density, and its organizing claim is that almost everything in that literature is a response to one fact: once a chemical potential $mu$ is turned on, the fermion determinant $det Delta(mu)$ becomes complex, so the importance-sampling step that makes lattice Monte Carlo work no longer applies. The paper's own contribution is a unification: it argues that the older density-of-states method already contains the recently proposed factorization method as a special case. Along the way it surveys the workarounds that existed in 2003 — Taylor expansion in $mu/T$, multi-parameter reweighting, imaginary chemical potential with analytic continuation, canonical-ensemble projection, strong-coupling expansions, complex Langevin — and it reports that two-color QCD, where the determinant stays real, shows quark-pair (diquark) condensation and a change in the vector-meson channel at finite density. A sympathetic reader should care because the cold, dense region of the QCD phase diagram is where color superconductivity and in-medium chiral restoration are expected, and this review is a claim about which computational routes into that region are actually open.

What carries the argument

Two pieces of machinery carry the argument. The first is the conjugation identity $det Delta(mu)^{*} = det Delta(-mu)$, derived from $Delta(mu)^{dagger} = gamma_5 Delta(-mu) gamma_5$; it is what makes the determinant real at zero density, real for SU(2) color because $U^{*} = sigma_2 U sigma_2$, and real for imaginary chemical potential, and it is also what makes the finite-isospin model ($mu_u = -mu_d$) simulable since its determinant is $|det Delta(mu)|^{2}$. The second is the Gibbs formula, $det Delta(mu) = e^{N_t V_s mu} prod_i (lambda_i + e^{-N_t mu})$, which condenses the determinant of a matrix of dimension $(3 N_x N_y N_z N_t)^{2}$ down to eigenvalues $lambda_i$ of a $mu$-independent matrix built from the spatial and mass terms; once those eigenvalues are known for a gauge configuration, the determinant can be evaluated at any chemical potential, which is what makes reweighting in $mu$ practical.

Extended reading notes

Core claim

The review's central claim is that the finite-density sign problem, not any single simulation, is the true subject of the field, and that the known methods can be sorted by how they dodge it. The authors state that the determinant obeys $(det Delta(mu))^{*} = det Delta(-mu)$, which makes it real at $mu = 0$, real for gauge group SU(2), and real for purely imaginary $mu$ — three escape routes that the review treats as the productive ones. On that basis it reports two substantive results from the literature: two-color QCD simulations, with a diquark source $j$ extrapolated to zero, indicate a low-temperature phase with nonzero $langle chi^{T} sigma_2 chi rangle$, and hadron propagators in that theory show the vector channel becoming lighter with density while pseudoscalar and diquark channels split apart. The second claim is methodological: the density-of-states method, in which the path integral is reduced to a one-dimensional integral over a histogrammed parameter weighted by the residual phase, is shown to contain the factorization method as an instance, so the two are not rival proposals but one idea in two costumes.

Load-bearing premise

The claim that two-color QCD condenses quark pairs rests on extrapolating an explicitly added diquark source to zero, and the review itself notes that at high density the small-source values fall away from the large-source trend and that the low-source discrepancy has been attributed to finite volume and a partially quenched approximation, with the thermodynamic limit still unresolved.

Editorial extensions

If this is right

  • If the sorting is right, then progress at high density and low temperature must come from algorithms that never sample a complex weight — imaginary-$mu$ continuation, canonical projection, density of states, or complex Langevin — rather than from better reweighting at real $mu$.
  • The two-color results imply that a confining gauge theory at nonzero density supports a color-singlet diquark condensate, which is superfluidity rather than color superconductivity, and therefore that the two phenomena should be distinguished experimentally and theoretically.
  • If the density-of-states method really contains the factorization method, then results obtained by the newer method can be cross-checked against the older literature, and the constrained-simulation machinery developed for the former transfers directly.
  • The observation that the vector-meson channel responds strongly to density while the pseudoscalar barely moves gives a concrete target for in-medium hadron searches, complementing the dilepton excesses reported by heavy-ion experiments.

Reading between the lines

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

  • The review's framing suggests a testable organizational prediction: methods should cluster by which symmetry of the determinant they exploit, so a method that works for SU(2) should fail for SU(3) exactly where the reality of the determinant is lost, and this could be checked by running the same algorithm on both groups at matched parameters.
  • Because the finite-isospin model shares the determinant modulus with the phase-quenched theory, any lattice result obtained under phase quenching is, strictly, a finite-isospin result; this makes the isospin phase diagram a calibration standard rather than a side topic.
  • The eigenvalue representation of the determinant suggests that the severity of the sign problem is controlled by the distribution of the $lambda_i$ relative to $e^{-N_t mu}$, which would let one predict in advance which $(mu, T)$ regions are reachable on a given lattice size without running the simulation.
  • The review treats complex Langevin as promising but unproven; the natural extension is that a kernel chosen to keep the Fokker-Planck spectrum's real part positive would be the decisive ingredient, and that this is checkable on the polynomial models before any QCD application.
Original abstract

This is a pedagogical review of the lattice study of finite density QCD, which is intended to provide the minimum necessary contents, so that the paper may be used as the first reading for a newcomer to the field and also for those working in nonlattice communities. After a brief introduction to argue why finite density QCD can be a new attractive subject, we describe fundamental formulae which are necessary for the following sections. Then we survey lattice QCD simulations in small chemical potential regions, where several prominent works have been reported recently. Next, two-color QCD calculations are discussed, where we have a chance to glance at many new features of finite density QCD, and indeed recent simulations indicated quark pair condensation and the in-medium effect. Tables of SU(3) and SU(2) lattice simulations at finite baryon density are given. In the next section, we make a survey of several related works which may be a starting point of the new development in the future, although some works do not attract much attention now. Materials are described in a pedagogical manner. Starting from a simple 2-d model, we briefly discuss a lattice analysis of NJL model. We describe a non-perturbative analytical approach, i.e., strong coupling approximation method and some results. Canonical ensemble approach instead of usual grand canonical ensample may be another route to reach high density. We examine the density of state method and show this old idea includes recently proposed factorization method. An alternative method, complex Langevin equation and an interesting model, finite isospin model, are also discussed. In Appendix, we give several technical points which are useful in practical calculations.

The compiled record

The paper's own sentences, arranged, kept beside the review and separate from it. Each label is drawn from a closed vocabulary; the number beside it is the probability the classifier gave that label. Numbers in the tables are the authors', checked by code. No sentence in this record was written by a model.

Compiled by classifier-1.13-20260917 via OpenRouter, question bank abstract-p1-20260923, 2026-09-23: 1 request, 9,439 input tokens, $0.0004 reported.

What the abstract itself states, as probabilities: claims first 0.03 · states limitation 0.35 · human participants 0.01 · controlled comparison 0.02 · randomized assignment 0.01 · states quantitative result 0.04 · compares with prior results 0.19 · mentions code or data availability 0.02.

The abstract, sentence by sentence

arXiv listing abstract v1. Results and the main claim are underlined; .

This is a pedagogical review of the lattice study of finite density QCD, which is intended to provide the minimum necessary contents, so that the paper may be used as the first reading for a newcomer to the field and also for those working in nonlattice communities. After a brief introduction to argue why finite density QCD can be a new attractive subject, we describe fundamental formulae which are necessary for the following sections. Then we survey lattice QCD simulations in small chemical potential regions, where several prominent works have been reported recently. Next, two-color QCD calculations are discussed, where we have a chance to glance at many new features of finite density QCD, and indeed recent simulations indicated quark pair condensation and the in-medium effect. Tables of SU(3) and SU(2) lattice simulations at finite baryon density are given. In the next section, we make a survey of several related works which may be a starting point of the new development in the future, although some works do not attract much attention now. Materials are described in a pedagogical manner. Starting from a simple 2-d model, we briefly discuss a lattice analysis of NJL model. We describe a non-perturbative analytical approach, i.e., strong coupling approximation method and some results. Canonical ensemble approach instead of usual grand canonical ensample may be another route to reach high density. We examine the density of state method and show this old idea includes recently proposed factorization method. An alternative method, complex Langevin equation and an interesting model, finite isospin model, are also discussed. In Appendix, we give several technical points which are useful in practical calculations.

background contribution method result interpretation main claim

What the classifier reports for each sentence. The number is the probability it gave the label.
#sentencerolethe authors' own workstates the main conclusionstrength
s0 This is a pedagogical review of the lattice study of finite density QCD, which is intended to provide the minimum necessary contents, so that the paper may be used as the first reading for a newcomer to the field and also for those working in nonlattice communities. contribution 0.81 0.17 0.05 asserted 0.90
s1 After a brief introduction to argue why finite density QCD can be a new attractive subject, we describe fundamental formulae which are necessary for the following sections. background 0.43 0.17 0.02 asserted 0.85
s2 Then we survey lattice QCD simulations in small chemical potential regions, where several prominent works have been reported recently. contribution 0.73 0.07 0.02 asserted 0.98
s3 Next, two-color QCD calculations are discussed, where we have a chance to glance at many new features of finite density QCD, and indeed recent simulations indicated quark pair condensation and the in-medium effect. contribution 0.63 0.15 0.07 hedged possibility 0.90
s4 Tables of SU(3) and SU(2) lattice simulations at finite baryon density are given. contribution 0.92 0.37 0.02 asserted 0.97
s5 In the next section, we make a survey of several related works which may be a starting point of the new development in the future, although some works do not attract much attention now. future work 0.61 0.12 0.03 hedged possibility 0.99
s6 Materials are described in a pedagogical manner. method 0.71 0.25 0.02 asserted 1.00
s7 Starting from a simple 2-d model, we briefly discuss a lattice analysis of NJL model. method 0.63 0.37 0.02 asserted 0.93
s8 We describe a non-perturbative analytical approach, i.e., strong coupling approximation method and some results. method 0.74 0.35 0.03 asserted 0.94
s9 Canonical ensemble approach instead of usual grand canonical ensample may be another route to reach high density. method 0.74 0.09 0.04 hedged possibility 1.00
s10 We examine the density of state method and show this old idea includes recently proposed factorization method. method 0.49 0.57 0.04 asserted 1.00
s11 An alternative method, complex Langevin equation and an interesting model, finite isospin model, are also discussed. method 0.66 0.16 0.03 asserted 0.86
s12 In Appendix, we give several technical points which are useful in practical calculations. contribution 0.81 0.47 0.02 asserted 0.97

What kind of paper the classifier reports this is: review 1.00 · evidence from literature 0.95.

What was examined, and what was not

sources read
  • arXiv listing abstract v1
not examined
  • full text
  • figures and tables
  • references and citing sentences
  • relations to other papers
levels run
  • P abstract
not run
  • S sections
  • T body sentences
  • C citations
  • R cross-paper relations

Every quoted sentence on this record is the authors'. Every label is from Pith's closed vocabulary, with the probability shown beside it. Every number in the measurements table is the authors', checked by code. Record as JSON

Figures
Figure 1. Schematic phase diagram of QCD in (T, µ) plane. 0 100 200 0 100 200 300 400 500 600 T [MeV] µ [MeV] Chiral Symmetry Broken <ψψ>=0 - / <ψψ>=0 CSC <ψψ>=0 - / <ψψ>=0 <ψψ>=0 - <ψψ>=0 precursor of CSC tricritical point
Figure 2. Phase diagram (T, µ) plane by NJL model. (Ref. 9)). Another interesting, but probably less well known possibility is chiral restoration in the nuclear medium. In the normal nuclear matter, the baryonic density is ρ0 ∼ 0.17/fm3 . In a Nambu-Jona-Lashino (NJL) type model, the quark condensation, hqq¯ i at ρ0 is less than that of the vacuum. We show the behavior of hψψ¯ i as a function of µ and T in a simple NJL model calculation in Fig.3. The quark condensation is an important parameter and causes many effects, such as vector meson masses. There are several recent experimental data…
Figure 3. hqq¯ i as a function of (T, µ). We solve gap equations at finite temperature and density derived in Ref. 20). We employ all parameter values given in 20), except tGD, which is set to be zero, i.e., we drop the s-quark contribution for simplicity. The arrow indicates the value of µ corresponding to the normal nuclear density ρ0. Hagedorn noted21) that if the hadron spectrum increases exponentially, ρ(m) = Ce m/T0 , (1.1) then the partition function diverges at some temperature Tc, called Hagedorn tem￾perature, Z dmρ(m)e−m/T = ∞ if T > T0. (1.2)
Figure 4. Phase diagram in the quark mass-chemical potential p
Figure 5. Eigen value distribution. Wilson fermions. 4
Figure 6. Eigen value distribution. Staggered fermions. 4
Figure 7. The triplet quark number susceptibility fo
Figure 8. The difference between singlet and triplet quark numb
Figure 9. The second derivative of the pseudo scalar mass
Figure 10. ∆P T 4 as a function of T /Tc for the values of µ/Tc. 3.2. Glasgow approach The Glasgow group has been developing the way towards simulations of lattice QCD at finite density. See Ref. 54) and references therein. Consider the partition function at µ normalized by that of µ = 0. Z = R DU det ∆(µ) e −βSG R DU det ∆(0) e−βSG = R DU det ∆(µ) det ∆(0) det ∆(0) e −βSG R DU det ∆(0) e−βSG =
Figure 11. Phase diagram obtained by Fodor and Katz. (Ref. 40))
Figure 12. Schematic diagram of multi-parameter reweighting
Figure 13. The phase of complex determinant calculated by Biel
Figure 14. Phase diagram obtained by Bielefeld-Swansea colla
Figure 15. The equation of state correction ∆(p/T 4 ) vs. T /Tc0 for various µ/Tc0 by Bielefeld-Swansea collaboration. (Ref. 42)). can be managed as the change of the link variables, Ut → zUt , which can be regarded as the simple change of the integration variables Ut to zUt . Therefore, the path integral gives the same result, Z(µI + φ) = Z(µI ). (3.30) 0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1 (aµΙ ) 2 5.28 5.29 5.3 5.31 5.32 5.33 5.34 β fit O(aµΙ ) 2 fit O(aµΙ ) 4 0 100 200 300 400 500 600 µB /MeV 160 165 170 175 180 T/MeV
Figure 16. Imaginary chemical potential study by de Forcrand a
Figure 17. The scaled chiral condensate, hψψ¯ i/hψψ¯ i0, (left) and number density, nB, (right) together with the chiral perturbation theory prediction. (Ref. 69)) It has been shown in Ref. 69) that the numerical data are well described by the chiral perturbation theory. Fig.17 presents the chiral condensate and the number density, respectively, together with the leading order prediction, hψψ¯ i hψψ¯ i0 =  1 1 x2 ;  mπ 8hψψ¯ i0  nB =  0 x < 1 x 4 (1 − 1 x4 ) x > 1 (4.7) where x ≡ 2µ mπ (4.8) Using extended QCD inequality, Kogut, Stephanov and Toublan have shown that in
Figure 18. Wilson line and condensate vs. β calculated by Kogut, Toublan and Sinclair. Standard gauge + Staggered fermions. NF = 4. Left : µ = 0.0 Right : µ = 0.3. (Ref. 63)) In Figs.18, we show the Wilson line, the chiral condensate hχχ¯ i and the diquark condensate hχσ2χi obtained by Kogut, Toublan and Sinclair.63) Here χ and ¯χ are staggered fermions. Using the staggered fermions, they simulated a system with a
Figure 19. Schematic phase diagram in the (T, µ) plane by Kogut, Toublan and Sinclair. (Ref. 63)) ∗) Without the source terms, the expectation values of diquark states, hχΓ χi, hχΓ¯ χ¯i vanish
Figure 20. Pseudo scalar and vector meson masses (left) and cor
Figure 21. Left : hσi as a function of µ on lattices of size 60 × Nτ ,Nτ = 6, 8, 10, 14, 20. Right : Coexisting states at µc. Ref.82). − 1 2a [ψ¯(x)(r − γ2)e −µaψ(x + ˆ2)ψ¯(x)(r + γ1)e µaψ(x − ˆ1)] + 1 a (2r + δm)ψ¯(x)ψ(x) − g 2 σ 2N (ψ¯(x)ψ(x))2 − g 2 π 2N (ψ¯(x)iγ5ψ(x))2 . (5.9) Here they introduce two couplings gσ and gπ to controle the continuum limit. The lattice effective potential is constructed with two auxiliary fields, σ and π as in the continuum, VL = −  δmL g 2 σ + 2C1  σL +  1 g 2 π − 2C0  π 2 L +  1 g 2 σ − C0 + 2C2  σ 2 L + 1
Figure 22. The diquark condensate, hqqi as a function of µ(left). hqqi as a function of j (right). potential, extrapolating to j → 0. From this analysis, they determine the existence of a nonzero BCS condensate at high baryon density (
Figure 23. Strong coupling calculations by Bili´c, Demeterfi a
Figure 24. (a) The chiral condensate hψψ¯ i = V −1 ∂logZ/∂(2ma) and baryon number density na3 = (3V Nt) −1 ∂logZ/∂(µat). (b) The average sign of the Boltzmann weights. (Ref. 91)) In Eq.(5.37), wK is not always positive. A sign σK for each monomer-dimer loop configuration is introduced as wK = σK|wK|, (5.38) and the simulation was carried out with the positive Boltzmann weight |wK|, hOi = hσKOi+ hσKi+ . (5.39) Fig.24 shows the chiral condensate and baryon number density as a function of µa for ma = 0.1 on 83 × 4 lattice obtained by Karsch and M¨utter. They determined that µca = 0.69± 0.01
Figure 25. Heavy quark potential for T ≃ 0.86Tc and B = 0 and 6. on 163 × 4. 5.5. Density of states method In lattice QCD, the average value of an operator hOi is given by hOi = 1 Z Z DUO[U] det ∆(U)e −βSg(U) , (5.43)
Figure 26. < ψψ > ¯ for nf = 4 on a 44 lattice at µI = 0.2 and mq = 0.025. The standard Monte Carlo results were obtained by the R-algorithm.103) Ref.105)
Figure 27. Finite Iso-spin phase diagram calculated by Son and
Figure 28. Numerical results of finite iso-spin model by Kogut a
Figure 29. Hadron propagators at µa = 0 (left) and µa = 0.8 (right). Pseudoscalar and vector mesons and scalar and axial vector baryons are shown. Figure (29) shows the propagators on a 43 × 8 lattice with spatially periodic boundary condition at β = 0.7 and κ = 0.150. At µ = 0 (29a), pseudoscalar meson and scalar di-quark, and vector boson and axial vector di-quark are degenerate. With the finite µ, the µB term in the W makes the di-quark propagators asymmetric in the time-direction. On the other hand, the meson propagators keep the symmetric shape. Therefore, degeneracy between the meson…

Figures reproduced from arXiv:hep-lat/0306031 by the authors.

References
How the literature uses it

62 arXiv papers cite this in 66 sentences

Each citing sentence is taken from the citing paper's LaTeX source and typed by Pith's classifier. A label appears at probability 0.8 or more (0.95 for support), a contest at 0.7 or more; the quoted words are the citing authors'. How each label was checked · JSON

Papers citing it · 36 sentences without a confident label

Uses its method or tool 1 paper
Cites it as background 28 papers

Discussion

Disagree with the review, or know something it missed? Leave a comment, highlight any passage above to comment on it, or challenge this review formally.