REVIEW 3 major objections 4 minor 27 references
Localization of strings on group manifolds
T0 review · 3 major / 4 minor · reviewed 2026-08-15 · deepseek-v4-flash
Pith's one-line read This paper proves that the WZW torus partition function can be computed exactly by supersymmetric localization as a sum over abelian winding modes, and verifies the result for SU(2) against the Weyl-Kac character formula.
desk verdict New localization formula for WZW partition functions with a real SU(2) check, held up by the contested level shift k -> k-2 and some deferred proofs. 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 supersymmetric localization of the type designed for theories in which the fermions are free and decoupled: one inserts the fermion zero-modes $ψ^{3}$_0 and \$tildeψ^{3}$_0 into the path integral so that the otherwise-vanishing supersymmetric index becomes nonzero, and deforms the action by a δ-exact term V = ∫ tr D_z \tildeψ (D_z \tilde J_z)^†. The bosonic part of δV is non-negative and vanishes only on classical solutions, which for the WZW model are abelian (maximal-torus-valued) maps labeled by winding numbers (m,w) together with a Weyl-group element ω. The one-loop determinant is evaluated by factorizing the fluctuation operator M = M1 M2, which produces the shift of the level from k to k-h; for SU(2), h=2. The Weyl-Kac character formula, the standard expression for affine-Lie-algebra characters as ratios of $\theta$ functions, is the comparison standard that the localized sum must match.
What would settle it
Evaluate the localized formula (3.57) for SU(3) and compare with the Weyl-Kac character sum at level k−3; a mismatch would show that the level shift k−h is not the universal prescription. Alternatively, compute the same one-loop determinant with a lattice regulator and check whether it equals the factorization-based determinant.
Extended reading notes
Core claim
The central claim of the paper is that the torus partition function of the supersymmetric WZW model can be computed exactly by supersymmetric localization, and that this determines the bosonic WZW partition function by division by the free-fermion partition function. Concretely, the localized supersymmetric partition function is a sum over Weyl-group elements ω and integer winding pairs (m,w): each term is exp(-$S^{{(ω)}}$_{m,w}) times a sign sgn(ω), where the classical action $S^{{(ω)}}$_{m,w} is that of an abelian solution embedded in the group via a maximal torus. The bosonic answer is obtained by dividing by Z_fer, the partition function of the decoupled adjoint fermions. The crucial check is that for G=SU(2) the localized formula (3.57), evaluated with the level shifted to k-2, is equal to the Weyl-Kac character sum (3.79) after a Poisson resummation. Thus the paper establishes that the WZW torus partition function is simultaneously a sum over quantum states (characters) and an exact sum over classical winding solutions.
Load-bearing premise
The whole agreement with the known answer depends on a particular way of handling the quantum fluctuation determinant, namely splitting it into two factors, which shifts the effective level from k to k−2 for SU(2); if that split is not the correct regularization, the localized sum will not reproduce the known partition function.
Editorial extensions
If this is right
- For SU(2), and by extension any compact simple simply-connected G, the WZW torus partition function has an exact representation as a sum over abelian winding sectors, complementing the usual sum over affine characters.
- The bosonic partition function is determined by the supersymmetric one through division by an explicit free-fermion factor, so the entire content of the bosonic model is encoded in the lattice sum.
- The localization method extends to noncompact and analytically continued models: SL(2,R) on a Lorentzian torus, where the result makes sense as a distribution, and the H3+/Z (BTZ black hole) model, whose partition function the paper computes and compares to earlier critical-point results.
- A Weyl-group sum with signs, plus a Wess-Zumino sign factor (−1)^ζ for general groups, keeps the localized formula modular invariant and gauge invariant in the diagonal and anti-diagonal twist sectors.
Reading between the lines
- A testable extension of the paper's logic: because fermion decoupling is local, the same zero-mode insertion should produce exact localization formulas for WZW correlation functions on higher-genus surfaces, not just the torus partition function.
- The paper's level shift k→k−h reads as a diagnostic: comparing the localization one-loop determinant with a lattice or Pauli-Villars regulator for G=SU(3) would show whether the shift k−h is the universal prescription or an artifact of the factorization M=M1M2.
- The near-agreement with the BTZ partition function, up to a factor sqrt((k−2)/k), suggests that localization might supply a first-principles derivation of the Euclidean black-hole partition function with a specific quantum correction to the thermal-circle volume; checking that factor microscopically would be a sharp test.
- One could reinterpret the Weyl-group sum and the sign sgn(ω) as a sum over conjugacy classes of the loop group, connecting the formula to standard index-theoretic results on group manifolds.
Signed reviews
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper presents a method, based on the Choi-Takhtajan supersymmetric-localization idea, for computing torus partition functions of bosonic WZW models. The strategy is to add free decoupled fermions, localize the supersymmetric model with an insertion of fermion zero-modes, and then divide by the free-fermion partition function to obtain the bosonic answer. The method is first illustrated for U(1), where it reproduces the standard free-boson result, and then applied in detail to SU(2), yielding a localized sum over abelian classical solutions labeled by winding numbers (m,w) and by Weyl-group elements. The claim is verified for SU(2) by a Poisson-resummation comparison with a Hamiltonian sum built from Weyl-Kac characters. The last sections extend the procedure to general compact groups and to analytic continuations such as SL(2,R) and H_3^+/Z, with the H_3^+/Z result compared to Maldacena-Ooguri-Son.
Significance. If the result held as stated, it would provide an exact classical-orbit sum formula for WZW partition functions, a non-trivial extension of the Choi-Takhtajan heat-kernel formula, and a new perspective on SL(2,R)/H_3^+ partition functions, including the BTZ black hole. The manuscript contains several concrete, checkable computations: the U(1) localization matches known results, the Poisson-summation identity (3.80)-(3.83) is explicit, and Appendix B gives a correct free-field identity for su(2)_1. These are genuine strengths. However, the central SU(2) verification is undermined by a load-bearing level-shift issue: the comparison is to characters at level k-2 rather than level k, and the paper's own Appendix B contradicts the level assignment used in Section 3.5. Because of this, the paper does not currently establish its main claim.
major comments (3)
- [Section 3.3, Eqs. (3.30)-(3.36)] The one-loop level shift k' = k - h is treated as an equally valid regularization choice, but the paper concedes that standard zeta-function and Pauli-Villars regularizations give k' = k. Since k is an integer, the statement that the two definitions differ only by a renormalization of k is not meaningful: there is no continuous renormalization connecting integer levels. This is load-bearing because the entire Weyl-Kac comparison uses characters at level k-2, not level k. A concrete test is k=1: the standard SU(2)_1 WZW model is a free boson with two characters, whereas the level-shifted formula (3.76) predicts k-1=0 characters. The manuscript's own Appendix B uses theta index 3 for su(2)_1, corresponding to the standard level k, not level k-2.
- [Section 3.5, Eqs. (3.73)-(3.79), and Appendix B] There is an internal contradiction in the level assignment. Equation (3.73) writes SU(2) characters at level k-2 using theta functions of index k, so for k=1 there are k-1=0 characters. Appendix B, however, correctly computes the su(2)_1 characters with theta index 3 and two characters, labeled by ell=1,2. Therefore the Poisson-resummation verification (3.80)-(3.83) proves only that the localized lattice sum equals a sum over level-(k-2) characters; it does not show equality with the partition function of the WZW model whose action is (3.3) with coefficient k. The advertised verification for G=SU(2) fails at k=1 and is conditional at all k.
- [Section 4, Eq. (4.2)] The general compact-group formula is presented as an expectation rather than a proven result, and the sign factor (-1)^zeta is said to have been determined by private communications and forthcoming papers [20,21], after being omitted in the original version. Since the abstract announces a computation for a general compact Lie group G, the unproven status of Eq. (4.2) and its sign factor is a substantive gap in the advertised generality, not a mere presentation issue.
minor comments (4)
- [Eq. (3.73)] The text refers to the affine extension u(2)_{k-2}; the intended object is su(2) at level k-2, and the notation should be corrected.
- [Eqs. (3.76) and (3.78)] The derivation of (3.78) from (3.76) via cancellation of theta functions assumes k >= 2; for k=1 the sum in (3.76) is empty while (3.78) is nonempty. The paper should state this limitation explicitly and address the k=1 case separately.
- [Section 5.3, Eq. (5.42)] The comparison with Maldacena-Ooguri-Son contains an undetermined factor sqrt(|k-2|/k), attributed to a possible quantum correction to the zero-mode volume. In an exact localization computation, an undetermined normalization weakens the check; this factor should either be computed or explicitly identified as an assumption.
- [References [19]-[21]] Several load-bearing statements, including the general-group formula and the sign factor, are supported by references listed as 'to appear' or as private communications. The manuscript should either include the necessary arguments or clearly state that the results depend on unpublished work.
Circularity Check
No circularity: the SU(2) matching is a nontrivial Poisson/theta identity benchmarked against Weyl-Kac, and the level-shift dependence is an explicitly flagged regularization caveat rather than a circular input.
full rationale
The claimed derivation is self-contained and not circular. The localization formula (3.57) is not obtained by assuming the Weyl-Kac answer; the one-loop prefactors are computed from explicit determinants and ground-state shifts (Appendix C), and the U(1)/S^1 result is independently checked against Polchinski's textbook expression (eq. 2.21). The SU(2) check is a genuine identity: after Poisson resummation, the localized lattice sum is shown equal to a theta-function sum built from Weyl-Kac characters (eqs. 3.80-3.83); no parameter has been fitted to force that equality. The paper does flag a substantive caveat in Sec. 3.3: the localization determinant, defined via the factorization M=M1M2, yields effective level k-h, whereas standard zeta/Pauli-Villars regularizations give k; the matching at eq. (3.76) therefore uses characters of u(2)_{k-2}, so the 'same partition function' comparison is conditional on that regularization choice. This is a correctness/regularization question, not a circularity, because the level shift is derived rather than chosen to match, and the paper explicitly juxtaposes the standard and factorization regularizations. Section 4's admission that an originally omitted sign (-1)^zeta was corrected by independent work [20,21] is likewise a limitation statement, not a circular reliance. The central result therefore does not reduce, by construction or by self-citation, to its own inputs.
Assumptions & free parameters
free parameters (1)
- Quantum correction to the h_0 zero-mode volume in H_3^+/Z =
undetermined; discrepancy factor sqrt(|k-2|/k) vs [26]
assumptions (5)
- domain assumption In the N=1 supersymmetric WZW model, the fermions are completely decoupled from the bosons by a chiral rotation, so the bosonic partition function is the supersymmetric one divided by the free-fermion partition function.
- domain assumption The localization data satisfy conditions (1.4)-(1.6): V and delta-V are independent of the absorbed fermion zero-modes, and the integral of the variation of the zero-mode product vanishes.
- ad hoc to paper The one-loop determinant is defined by factoring M = M1 M2 (eqns (3.34)-(3.36)) and this is the physically correct regularization for the bosonic WZW model; it shifts the effective level to k' = k - h, differing from standard regularizations (which give k' = k) only by a renormalization of k.
- ad hoc to paper For general compact G, the localized formula (4.2), including the sign factor (-1)^zeta, equals the Hamiltonian sum over Weyl-Kac characters.
- standard math Standard mathematical background: holomorphic line bundles of degree zero on a torus, the Weyl-Kac character formula, Poisson summation, Jacobi theta identities, and the distribution interpretation of theta functions on the real axis (proposition 4.2 of [25]).
Cite this review
Pith. "Pith review of Localization of strings on group manifolds." pith.science (2026). https://pith.science/paper/442REJ5N
@misc{pith2026250620028,
author = {Pith},
title = {Pith review of: Localization of strings on group manifolds},
year = {2026},
howpublished = {\url{https://pith.science/paper/442REJ5N}},
note = {Machine review of arXiv:2506.20028}
}
abstract
We compute the partition function of the WZW model with target a compact Lie group $G$ by adapting a method used by Choi and Takhtajan to compute the heat kernel of the group manifold. The basic idea is to compute the partition function of a supersymmetric version of the WZW model using a form of supersymmetric localization and then use the fact that, since the fermions of the supersymmetric WZW model are actually decoupled from the bosons, this also determines the partition function of the purely bosonic WZW model. The result is a formula for the partition function as a sum over contributions from abelian classical solutions. We verify for $G=SU(2)$ that this formula agrees with the result for the same partition function that comes from the Weyl-Kac character formula. We extend the method of supersymmetric localization to certain related models such as the $SL(2,\mathbb{R})$ WZW model and a Wick-rotated version of this model in which the target space is hyperbolic three-space $H_3^+$.
Reference graph
Works this paper leans on
-
[1]
C. Choi and L. A. Takhtajan,Supersymmetry and trace formulas. Part I. Compact Lie groups,JHEP06(2024) 026, [2112.07942]
arXiv 2024
-
[2]
P. Di Vecchia, V. G. Knizhnik, J. L. Petersen and P. Rossi,A Supersymmetric Wess-Zumino Lagrangian in Two-Dimensions,Nucl. Phys. B253(1985) 701–726
work page 1985
-
[3]
L. D. `Eskin,Heat equation on Lie groups, inIn Memoriam: N. G. Cebotarev (Russian), pp. 113–132. Izdat. Kazan. Univ., Kazan ′, 1964
work page 1964
-
[4]
Supersymmetry and trace formulas III. Frenkel trace formula
C. Choi and L. A. Takhtajan,Supersymmetry and trace formulas III. Frenkel trace formula, 2502.10210
-
[5]
S. Cecotti, P. Fendley, K. A. Intriligator and C. Vafa,A New supersymmetric index,Nucl. Phys. B386(1992) 405–452, [hep-th/9204102]
arXiv 1992
-
[6]
J. M. Maldacena, G. W. Moore and A. Strominger,Counting BPS black holes in toroidal Type II string theory,hep-th/9903163
-
[7]
Kiritsis,Introduction to nonperturbative string theory,AIP Conf
E. Kiritsis,Introduction to nonperturbative string theory,AIP Conf. Proc.419(1998) 265–308, [hep-th/9708130]
arXiv 1998
-
[8]
Witten,Two-dimensional gauge theories revisited,J
E. Witten,Two-dimensional gauge theories revisited,J. Geom. Phys.9(1992) 303–368, [hep-th/9204083]
arXiv 1992
Show all 27 references
-
[9]
Polchinski,String theory
J. Polchinski,String theory. Vol. 2: Superstring theory and beyond. Cambridge Monographs on Mathematical Physics. Cambridge University Press, 12, 2007. 10.1017/CBO9780511618123
2007 doi
-
[10]
Polchinski,String theory
J. Polchinski,String theory. Vol. 1: An introduction to the bosonic string. Cambridge Monographs on Mathematical Physics. Cambridge University Press, 12, 2007. 10.1017/CBO9780511816079
2007 doi
-
[11]
Witten,Nonabelian Bosonization in Two-Dimensions,Commun
E. Witten,Nonabelian Bosonization in Two-Dimensions,Commun. Math. Phys.92(1984) 455–472
1984
-
[12]
Gawedzki and A
K. Gawedzki and A. Kupiainen,G/h Conformal Field Theory from Gauged WZW Model, Phys. Lett. B215(1988) 119–123
1988
-
[13]
Gawedzki and A
K. Gawedzki and A. Kupiainen,Coset Construction from Functional Integrals,Nucl. Phys. B 320(1989) 625–668
1989
-
[14]
Witten,On Holomorphic factorization of WZW and coset models,Commun
E. Witten,On Holomorphic factorization of WZW and coset models,Commun. Math. Phys. 144(1992) 189–212
1992
-
[15]
Dei and E
A. Dei and E. J. Martinec,NS5-brane backgrounds and coset CFT partition functions,JHEP 06(2024) 147, [2403.17258]
2024 arXiv
-
[16]
Witten,The N matrix model and gauged WZW models,Nucl
E. Witten,The N matrix model and gauged WZW models,Nucl. Phys. B371(1992) 191–245
1992
-
[17]
Henningson,N=2 gauged WZW models and the elliptic genus,Nucl
M. Henningson,N=2 gauged WZW models and the elliptic genus,Nucl. Phys. B413(1994) 73–83, [hep-th/9307040]
1994 arXiv
-
[18]
Jackiw and C
R. Jackiw and C. Rebbi,Solitons with Fermion Number 1/2,Phys. Rev. D13(1976) 3398–3409
1976
-
[19]
To appear
S. Murthy and D. Zagier, “To appear.” – 57 –
-
[20]
Zhao,WZW Partition Functions from Supersymmetric Localization,2507.11673
B. Zhao,WZW Partition Functions from Supersymmetric Localization,2507.11673
-
[21]
Private communication, and to appear
Y. L¨ u, “Private communication, and to appear.”
-
[22]
J. M. Maldacena and H. Ooguri,Strings in AdS(3) and SL(2,R) WZW model 1.: The Spectrum,J. Math. Phys.42(2001) 2929–2960, [hep-th/0001053]
2001 arXiv
-
[23]
Choi and L
C. Choi and L. A. Takhtajan,Supersymmetry and trace formulas II. Selberg trace formula, 2306.13636
-
[24]
S. K. Ashok and J. Troost,Path integrals on sl(2, R) orbits,J. Phys. A55(2022) 335202, [2204.00232]
2022 arXiv
-
[25]
Bostelmann and C
H. Bostelmann and C. J. Fewster,Quantum inequalities from operator product expansions, Communications in Mathematical Physics292(July, 2009) 761–795, [0812.4760]
2009 arXiv
-
[26]
J. M. Maldacena, H. Ooguri and J. Son,Strings in AdS(3) and the SL(2,R) WZW model. Part 2. Euclidean black hole,J. Math. Phys.42(2001) 2961–2977, [hep-th/0005183]
2001 arXiv
-
[27]
Gawedzki,Noncompact WZW conformal field theories, inNATO Advanced Study Institute: New Symmetry Principles in Quantum Field Theory, pp
K. Gawedzki,Noncompact WZW conformal field theories, inNATO Advanced Study Institute: New Symmetry Principles in Quantum Field Theory, pp. 0247–274, 10, 1991. hep-th/9110076. – 58 –
1991 arXiv
Reviewed August 15, 2026 · model on record in the stance chip above.
Discussion (0). Continue with ORCID to comment.