Pith. sign in

REVIEW 1 major objections 6 minor 1 cited by

The canonical generalised Levi-Civita connection and its curvature

T0 review · 1 major / 6 minor · reviewed 2026-08-06 · deepseek-v4-flash

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

Pith's one-line read For every generalized metric and divergence operator on an exact Courant algebroid, this paper constructs a canonical generalized Levi-Civita connection whose curvature is an invariant of the pair.

desk verdict A careful, genuinely useful canonical construction that resolves non-uniqueness of generalized Levi-Civita connections; the key nondegeneracy lemma is compressed but correct, and the paper deserves refereeing. read the letter →

arxiv 2507.17604 v1 pith:LKO7V3BP submitted 2025-07-23 math.DG gr-qchep-th

Vicente Cortés , Matas Mackevicius , Thomas Mohaupt , Oskar Schiller This is my paper · ORCID
classification math.DGgr-qchep-th MSC 53D1883C10
keywords generalisedgeometryCourantalgebroidsLevi-CivitaconnectionmetricdivergenceoperatorRiemanntensorEinsteinequationssupergravity
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

This paper solves a well-known ambiguity in generalized geometry: generalized Levi-Civita connections are not unique, so curvature quantities built from them depend on a choice. The authors construct, for each exact Courant algebroid equipped with a (semi-Riemannian) generalized metric $\mathcal G$ and a divergence operator $\mathrm{div}$, a single canonical torsion-free metric connection $D^{\mathcal G,\mathrm{div}}$. Its generalized Riemann tensor is therefore an invariant of the pair $(\mathcal G,\mathrm{div})$, not of an arbitrary choice. The paper's main formulas express that tensor, and the associated full Ricci tensor, Ricci tensor, and three scalar invariants, entirely in terms of ordinary Riemann curvature, the closed three-form $H$ of the Courant algebroid, and the section $e = 2(X+\xi)$ that records the difference between $\mathrm{div}$ and the metric divergence. A reader should care because these master formulas make generalized geometry directly usable in gravitational and string-theoretic settings, recovering standard supergravity equations and extending them to vector-deformed supergravity.

What carries the argument

The load-bearing construction is the trace-map orthogonal projection inside the first generalized prolongation, the space of tensors $A\in (\mathrm{so}(E)_{\mathcal G})^{\langle 1\rangle}\subset E^*\otimes \Lambda^2 E^*$ satisfying the algebraic condition that the cyclic sum $\partial A$ vanishes. For a pair $(\mathcal G,\mathrm{div})$, the canonical connection is $D^{\mathcal G,\mathrm{div}}=D^0+S$, where $S$ is the unique section of this prolongation perpendicular to the kernel of the trace map $S\mapsto \mathrm{tr}(S)$ and with prescribed trace $\alpha=\mathrm{div}-\mathrm{div}^{\mathcal G}$; Lemma 2.2 supplies the nondegeneracy of the kernel that makes $S$ well-defined. The curvature computation then reduces to using the closed form $S=(\chi^{e_+}_+ + \chi^{e_-}_-)/(d-1)$ and expanding $\mathrm{Rm}^{D^0+S}$ into $D^0$-derivatives of $S$ plus algebraic terms quadratic in $S$, producing the master formulas of Theorems 4.1 and 4.2.

What would settle it

Compute the Gram matrix of the kernel of $\mathrm{tr}: \mathrm{so}(V)^{\langle 1\rangle}\to V^*$ on a pseudo-Euclidean space with a mixed-signature metric (positive and negative definite parts both nonzero, e.g. signature $(2,2)$) and check whether it is degenerate; a degenerate kernel would exhibit a nonzero $S$ with $\mathrm{tr}(S)=0$ but $\langle S,S\rangle=0$, breaking the claimed uniqueness of $S_\alpha$ and invalidating $D^{\mathcal G,\mathrm{div}}$ for arbitrary $\mathrm{div}$. The paper's own proof asserts the kernel is a direct sum of two definite subspaces, so this is a direct numerical test.

Watch

Extended reading notes

Core claim

The central claim is that for any exact Courant algebroid with generalized metric $\mathcal G$ and divergence operator $\mathrm{div}$, there exists a canonical generalized Levi-Civita connection $D^{\mathcal G,\mathrm{div}}$, obtained as $D^0+S$, where $D^0$ is the extension of the Levi-Civita connection of the induced metric with metric divergence, and $S$ is the unique section of the first generalized prolongation $\mathrm{so}(E)^{\langle 1\rangle}_{\mathcal G}$ that has trace $\mathrm{div}-\mathrm{div}^{\mathcal G}$ and is orthogonal to the trace-free tensors. Its generalized Riemann tensor $\mathrm{Rm}^{D}$ decomposes into pure-type and mixed-type components expressed in closed form by Theorems 4.1 and 4.2 in terms of the ordinary Riemann tensor, the closed three-form $H$, and the section $e$ with $\langle e,\cdot\rangle=\mathrm{div}^{\mathcal G}-\mathrm{div}$. Tracing these formulas yields the full generalized Ricci tensor, the generalized Ricci tensor, and three scalar invariants, two of which are new (the second generalized scalar curvature and the generalized Kretschmann scalar). In the metric-divergence case $e=0$ the formulas recover a known class of generalized Levi-Civita connections from the literature, and in the dilaton and vector-deformed cases they match, respectively, the NS-NS supergravity equations and the field equations of vector-deformed supergravity.

Load-bearing premise

The whole construction stands on the algebraic claim, proved in Lemma 2.2, that the kernel of the trace map on the first generalized prolongation $\mathrm{so}(E)^{\langle 1\rangle}_{\mathcal G}$ is nondegenerate; if that kernel were degenerate for some signature or dimension, the unique section $S$ and hence the canonical connection would not exist for arbitrary divergence operators.

Editorial extensions

If this is right

  • Any invariant computed from $D^{\mathcal G,\mathrm{div}}$, in particular its generalized Riemann tensor and the new scalar invariants, is a genuine invariant of the pair $(\mathcal G,\mathrm{div})$; no curvature component has to be discarded for depending on a connection choice.
  • The master formulas reduce generalized curvature computations to classical data: ordinary Riemann curvature, the closed three-form $H$, and the section $e$; one no longer needs to track the full generalized connection.
  • Tracing the formulas produces explicit expressions for the full generalized Ricci tensor, the generalized Ricci tensor, and three scalar invariants, ready for use in generalized Einstein equations.
  • Under the standard dilaton identification $\xi=2\,d\varphi$, $X=0$, the formulas reproduce the NS-NS supergravity field equations; under the compatibility conditions of vector-deformed supergravity, vanishing of the second generalized scalar curvature is part of the field equations.
  • The generalized Kretschmann scalar $|\mathrm{Rm}^{D}|^2_{\mathcal G}$ is a new invariant that in the flat case $H=0$, $e=0$ reduces to $4|\mathrm{Rm}|^2_g$, so it can serve as a gravitational singularity probe in generalized geometry.

Reading between the lines

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

  • The canonical connection gives a preferred representative of the equivalence class of generalized Levi-Civita connections with fixed divergence, which may make generalized Ricci flow and generalized Einstein equations into well-posed evolution problems rather than gauge-choice-dependent ones.
  • The explicit dependence on $e$ suggests a systematic classification of flat pairs $(\mathcal G,\mathrm{div})$: solving $\mathrm{Rm}^{D}=0$ as equations for $g$, $H$, and the vector field $X$ and one-form $\xi$ could characterize exact Courant algebroids admitting Ricci-flat generalized geometry.
  • The new scalar invariants, particularly the generalized Kretschmann scalar, invite applications to curvature singularity and extendability questions in generalized geometry, a direction the paper notes only in passing.
  • Because the construction works for semi-Riemannian metrics of any signature, it may be directly relevant to timelike T-duality and other signature-changing settings in gravitational theories.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

1 major / 6 minor

Summary. The paper constructs, for an exact Courant algebroid with a semi-Riemannian generalized metric G and a divergence operator div, a canonical generalized Levi-Civita connection D^{G,div}, thereby resolving the non-uniqueness of torsion-free metric generalized connections. It then computes the generalized Riemann tensor of D^{G,div} and decomposes it into classical data: the ordinary Riemann tensor of the underlying metric, the closed three-form H, and a section e encoding the divergence difference. The main curvature formulas appear in Theorems 4.1 and 4.2 (pure- and mixed-type components), with corollaries for the full generalized Ricci tensor, the generalized Ricci tensor, two scalar curvatures, and a newly defined generalized Kretschmann scalar. Section 5 compares the results with the physics literature, recovering known expressions for NS-NS supergravity and vector-deformed supergravity.

Significance. If the results are correct, this is a substantial contribution to generalized geometry: it turns the non-uniqueness of generalized Levi-Civita connections into a canonical choice, makes the generalized Riemann tensor an invariant of the pair (G, div), and provides explicit master formulas that can be used in applications. The curvature computations are carried out in detail and are benchmarked against the existing literature [1, 11, 14], which gives strong independent consistency checks. The new scalar-valued invariants, including the generalized Kretschmann scalar, are potentially useful for future work on curvature invariants and supergravity applications. The main caveat is the compressed proof of Lemma 2.2, which is load-bearing for the existence and uniqueness of the canonical connection; the algebraic argument is sketched rather than fully demonstrated. This is a presentational gap in an otherwise sound construction.

major comments (1)
  1. [Lemma 2.2, §2] The proof of nondegeneracy of the kernel of the trace map tr: so(E)^{⟨1⟩}_G → E* is too compressed at the key step. The sentence "It clear that the equation ∂S = 0 decouples into four independent equations corresponding to the tensor power of L" is not a proof: one must explicitly show that the four components of ∂S with different L-degree vanish independently, and that the subspaces so(V)^{⟨1⟩}_{odd} and so(V)^{⟨1⟩}_{even} are definite with respect to the relevant inner product. In addition, the splitting ker(tr: so(V)^{⟨1⟩}_{odd} → L*) ⊕ ker(tr: so(V)^{⟨1⟩}_{even} → P*) requires an explicit verification that the trace of an odd-parity element takes values in L* and that of an even-parity element in P*. Because the existence of the complementary subbundle K = (ker tr)⊥ and the unique section S_α with prescribed trace depend on this lemma, the proof must be completed and presented in detail.
minor comments (6)
  1. [§2, first paragraph of proof of Theorem 2.1] The notation "TM = TM ⊕ T*M" is confusing because the symbol TM is used both for the generalized tangent bundle and for the ordinary tangent bundle. Please use a distinct notation, e.g. \mathbb{T}M or E, for the generalized tangent bundle.
  2. [Abstract and header] The classification line reads "MSc classification"; this should be "MSC classification".
  3. [Introduction, page 6] There is a typo "connnection" in the sentence "The Riemann tensor of the canonical generalised Levi-Civita connnection".
  4. [Lemma 2.2, proof] The phrase "It clear that the equation ∂S = 0 decouples" should read "It is clear that the equation ∂S = 0 decouples".
  5. [Section 5, index notation for the generalized Ricci tensor] In the expression for 4Rc(G, div)±_μν, the meaning of the ± and ∓ signs (one for each chirality) should be stated explicitly, since the left-hand side contains a superscript ± that is not defined in the text.
  6. [Proposition 4.6] The non-triviality check (|Rm^D|^2_G)|_{H=0,e=0} = 4|Rm|^2_g is stated without derivation. A short justification, or a reference to the relevant formulas, would be helpful.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the canonical connection and its curvature are derived from explicit constructions and external benchmarks; the only self-citation is a definitional convention, not load-bearing.

full rationale

The paper's central construction is self-contained. Theorem 2.1 defines D0 explicitly from the Levi-Civita connection of the underlying metric and then defines D^{G,div} = D0 + S, with S uniquely determined by Lemma 2.2, whose proof is carried out in the paper. The curvature formulas in Theorems 3.5, 3.6, 4.1, and 4.2 are derived from the definition (3.1) by direct computation, and they are benchmarked against the independent references [1], [11], and [14]. No parameter is fitted to data and then renamed a prediction; the field e = 2(X+ξ) is introduced as divG - div and the formulas express dependence on this input, which is not circular. The only same-author citation entering the main definition is (3.1), taken from [2]; this is a stipulated definition of the generalized Riemann tensor, not an unverified theorem, and the subsequent computations do not presuppose the target curvature decompositions. The comparisons in Section 5 are consistency checks, not circular validations: specializing e to the dilaton case recovers the known NS-NS formulas of [11], and specializing to the vector-deformed case matches [10]. Thus no load-bearing circular step was identified.

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

No fitted numerical parameters appear; G and div are input data. The assumptions are the standard structures of exact Courant algebroids plus the specific conventions imported from [2] and [13]. The main structural risk is the nondegeneracy of the trace-map kernel in Lemma 2.2, which is needed for the canonical complement.

assumptions (4)
  • domain assumption Exact Courant algebroid structure with anchor pi and closed three-form H; a semi-Riemannian generalized metric G determines isomorphisms pi_plus, pi_minus from E_plus, E_minus to TM.
    This is the setting of Section 2; it fixes the underlying metric g and the identifications used in all curvature formulas.
  • domain assumption The generalized Riemann tensor is defined by equation (3.1), as recalled from the companion preprint [2] by two of the same authors.
    The central invariant is this specific convention; the paper does not re-derive it. Comparisons with [14] and [15] mitigate the self-citation burden.
  • domain assumption The first generalized prolongation so(E)^{<1>}_G and its trace map are used as in [13] by Cortes and David.
    Lemma 2.2 and Proposition 3.4 rely on this framework; the details are not reproduced in the paper.
  • standard math Standard Bianchi identities and trace conventions for Rm^D are assumed to identify the curvature tensor by its pure- and mixed-type components.
    Used in Section 3 before Theorem 3.5; standard practice in generalized geometry.

how reviews work

0 comments
Cite this review

Pith. "Pith review of The canonical generalised Levi-Civita connection and its curvature." pith.science (2026). https://pith.science/paper/LKO7V3BP

@misc{pith2026250717604,
  author       = {Pith},
  title        = {Pith review of: The canonical generalised Levi-Civita connection and its curvature},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/LKO7V3BP}},
  note         = {Machine review of arXiv:2507.17604}
}
abstract

Given a (semi-Riemannian) generalised metric $\mathcal G$ and a divergence operator $\mathrm{div}$ on an exact Courant algebroid $E$, we geometrically construct a canonical generalised Levi-Civita connection $D^{\mathcal G, \mathrm{div}}$ for these data. In this way we provide a resolution of the problem of non-uniqueness of generalised Levi-Civita connections. Since the generalised Riemann tensor of $D^{\mathcal G, \mathrm{div}}$ is an invariant of the pair $(\mathcal G, \mathrm{div})$, we no longer need to discard curvature components which depend on the choice of the generalised connection. As a main result we decompose the generalised Riemann curvature tensor of $D^{\mathcal G, \mathrm{div}}$ in terms of classical (non-generalised) geometric data. Based on this set of master formulas we derive a comprehensive curvature tool-kit for applications in generalised geometry. This includes decompositions for the full generalised Ricci tensor, the generalised Ricci tensor, and three generalised scalar-valued curvature invariants, two of which are new.

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. Gauged Extended Field Theory and Generalised Cartan Geometry

    hep-th 2025-09 conditional novelty 6.0 of 10

    A systematic Cartan-geometric construction of linearised torsion and curvature hierarchies for generalised geometries with global duality group G and local gauge group H, realised via brane current algebras.

Reference graph

Works this paper leans on

17 extracted references · 9 canonical work pages · cited by 1 Pith paper

  1. [1]

    Springer,

    Mario Garcia-Fernandez and Jeffrey Streets.Generalized Ricci Flow. Springer,

    work page

  2. [11]

    Super- gravity as Generalised Geometry I: Type II Theories.JHEP, 11:091, 2011

    Andre Coimbra, Charles Strickland-Constable, and Daniel Waldram. Super- gravity as Generalised Geometry I: Type II Theories.JHEP, 11:091, 2011. doi:10.1007/JHEP11(2011)091

    doi:10.1007/jhep11(2011)091 2011

  3. [10]

    Arutyunov, S

    G. Arutyunov, S. Frolov, B. Hoare, R. Roiban, and A. A. Tseytlin. Scale invariance of theη-deformedAdS5×S5 superstring, T-duality and modified type II equations. Nucl. Phys. B, 903:262–303, 2016. doi:10.1016/j.nuclphysb. 2015.12.012. 34 References

    doi:10.1016/j.nuclphysb 2016

  4. [2]

    Exterior Generalised Geometry

    Vicente Cortés and Oskar Schiller. Exterior generalised geometry. doi:10. 48550/arXiv.2507.12362

    work page Pith review arXiv doi:10.48550/arxiv.2507.12362

  5. [3]

    T. H. Buscher. A Symmetry of the String Background Field Equations.Phys. Lett. B, 194:59–62, 1987. doi:10.1016/0370-2693(87)90769-6

    doi:10.1016/0370-2693(87)90769-6 1987

  6. [4]

    Target space duality in string theory.Phys

    Amit Giveon, Massimo Porrati, and Eliezer Rabinovici. Target space duality in string theory.Phys. Rept., 244:77–202, 1994. doi:10.1016/0370-1573(94) 90070-1

    doi:10.1016/0370-1573(94 1994

  7. [5]

    Cavalcanti and Marco Gualtieri

    Gil R. Cavalcanti and Marco Gualtieri. Generalized complex geometry and T-duality. InA Celebration of the Mathematical Legacy of Raoul Bott, CRM Proceedings and Lecture Notes, pages 341–366. American Mathematical Society,

    work page

  8. [6]

    Transitive Courant Algebroids, String Structures and T-duality

    David Baraglia and Pedram Hekmati. Transitive Courant Algebroids, String Structures and T-duality. Adv. Theor. Math. Phys., 19:613–672, 2015. doi: 10.4310/ATMP.2015.v19.n3.a3

    doi:10.4310/atmp.2015.v19.n3.a3 2015

Show all 17 references
  1. [7]

    T-duality for transitive courant algebroids.J

    Vicente Cortés and Liana David. T-duality for transitive courant algebroids.J. Symplectic Geom., 21(4):775–856, 2023. doi:10.4310/JSG.2023.v21.n4.a4

    2023 doi

  2. [8]

    T-duality across non-extremal horizons

    Maxime Medevielle and Thomas Mohaupt. T-duality across non-extremal horizons. JHEP, 09:116, 2024. doi:10.1007/JHEP09(2024)116

    2024 doi

  3. [9]

    Higher derivative supergravities in diverse dimensions.Phys

    Mehmet Ozkan, Yi Pang, and Ergin Sezgin. Higher derivative supergravities in diverse dimensions.Phys. Rept., 1086:1–95, 2024. doi:10.1016/j.physrep. 2024.07.002

    2024 doi

  4. [12]

    Classification of generalized einstein metrics on 3-dimensional lie groups

    Vicente Cortés and David Krusche. Classification of generalized einstein metrics on 3-dimensional lie groups. Canad. J. Math., 75(6):2038–2095, 2023. doi: 10.4153/S0008414X23000056

    2023 doi

  5. [13]

    Vicente Cortés and Liana David. Generalized connections, spinors, and inte- grability of generalized structures on courant algebroids.Moscow Mathematical Journal, 21(4):695–736, 2021.doi:10.17323/1609-4514-2021-21-4-695-736

    2021 doi

  6. [14]

    Ricci flow on courant algebroids.Communications in Contemporary Mathematics, page 2550037, 2025

    Jeffrey Streets, Charles Strickland-Constable, and Fridrich Valach. Ricci flow on courant algebroids.Communications in Contemporary Mathematics, page 2550037, 2025. doi:10.1142/S0219199725500373

    2025 doi

  7. [15]

    Cavalcanti, Jaime Pedregal, and Roberto Rubio

    Gil R. Cavalcanti, Jaime Pedregal, and Roberto Rubio. On the equivalence of generalized Ricci curvatures.Proc. Amer. Math. Soc., 153(6):2639–2648, 2025. doi:10.1090/proc/17165. V. Cortés: vicente.cortes@uni-hamburg.de Department of Mathematics, University of Hamburg, Bundesstr...

    2025 doi

  8. [2010]

    doi:10.1090/crmp/050

    doi

  9. [2020]

    doi:10.1090/ulect/076

    doi

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.