Description of curved spacetimes by finite-size matrices in the type IIB matrix model

Abstract: The type IIB matrix model is expected to give a nonperturbative formulation of superstring theory. Its covariant derivative interpretation provides a method to describe curved spacetimes in the model. There, matrices are identified with certain covariant derivatives which can be viewed as infinite-size matrices. Here, by using the Berezin-Toeplitz quantization, we develop a method to regularize these matrices as finite-size ones, which is needed to calculate quantum effects in the interpretation or in particular to apply the interpretation to the results of numerical simulations. As examples, we examine the cases of $T<sup>{2n}$ and $S<sup>2$ in detail.

• The paper constructs Hermitian finite-size matrices representing covariant derivatives through Berezin–Toeplitz quantization, with convergence to the continuum operator at O(1/p).
• The regularized derivative commutators reproduce bundle-curvature algebra, while explicit tests on flat tori and S² recover commuting derivatives and large-p SU(2) closure.
• The method extends curved-spacetime studies in the type IIB matrix model to quantizable closed Kähler manifolds, although embedding dependence, Lorentzian geometries, and numerical applications remain open issues.

Motivation and setting

The type IIB matrix model, obtained by dimensional reduction of ten-dimensional $U(N)$ $\mathcal{N}=1$ super Yang-Mills theory to zero dimensions, is a candidate nonperturbative definition of superstring theory in which spacetime is expected to emerge dynamically from the eigenvalue structure of ten Hermitian matrices $A_a$ (2606.24577). A central requirement for such a formulation is the ability to describe curved spacetimes. The covariant derivative interpretation of Hanada, Kawai, and Kimura provides one established framework: the matrices are identified with covariant derivatives $\nabla_{(a)}$ acting on sections of a bundle whose fiber carries the regular representation of $G = \mathrm{Spin}(d)$ or $\mathrm{Spin}^c(d)$. In this interpretation, Einstein's equations with cosmological constant $\Lambda = -(d-2)m^2$ follow from the mass-deformed equations of motion, and general coordinate and local Lorentz transformations are embedded in the model's $U(N)$ symmetry.

The obstruction to quantum calculations is that each component $$\nabla_{(a)} = R^{\langle v\rangle}_{(a)b}(g^{-1})\,\nabla_b$$ acts as an endomorphism on an infinite-dimensional space $\Gamma(E_{\mathrm{reg}})$, so the corresponding "matrices" are infinite-dimensional operators. Since extracting geometry from Monte Carlo data on the type IIB matrix model requires finite-size matrices, a systematic regularization is needed. This paper constructs such a regularization via Berezin-Toeplitz (BT) quantization, generalizing earlier work on closed Riemann surfaces (Hattori et al., 2024) to closed connected Kähler manifolds of real dimension $\mathcal{N}=1$0.

Construction via Berezin-Toeplitz quantization

The BT quantization regularizes a section $\mathcal{N}=1$1 as a Toeplitz operator $\mathcal{N}=1$2, where $\mathcal{N}=1$3 projects onto the finite-dimensional kernel of the Dirac operator twisted by $\mathcal{N}=1$4. For sufficiently large topological charge $\mathcal{N}=1$5, the vanishing theorem and Atiyah-Singer index theorem guarantee that $\mathcal{N}=1$6 with $\mathcal{N}=1$7, so the zero-mode space grows as $\mathcal{N}=1$8. In this semiclassical limit, products of Toeplitz operators admit an asymptotic star-product expansion, yielding the correspondences $\mathcal{N}=1$9 and $A_a$0.

The paper's central result is the finite-size matrix $A_a$1 representing $A_a$2, defined through its action on the Toeplitz operator $A_a$3 of a section $A_a$4:

$A_a$5

where $A_a$6 ($A_a$7) are isometric embedding coordinates of $A_a$8 into $A_a$9 (existence guaranteed by Nash's theorem), and the bracket $\nabla_{(a)}$0 accounts for the rectangular shape of $\nabla_{(a)}$1, which arises because left- and right-acting Toeplitz operators have different sizes.

Two structural features of the construction deserve emphasis. First, since the regular representation decomposes into infinitely many irreducibles, the Peter-Weyl expansion of $\nabla_{(a)}$2 must be truncated by a cutoff $\nabla_{(a)}$3 on the Casimir of the irreducible representations; the physical limit is taken as $\nabla_{(a)}$4 and $\nabla_{(a)}$5 with $\nabla_{(a)}$6. Second, the second term in the definition of $\nabla_{(a)}$7, though subleading at $\nabla_{(a)}$8, is essential: it renders $\nabla_{(a)}$9 Hermitian with respect to the Frobenius inner product for finite $G = \mathrm{Spin}(d)$0, which the authors prove explicitly using the Hermiticity of $G = \mathrm{Spin}(d)$1, $G = \mathrm{Spin}(d)$2, and $G = \mathrm{Spin}(d)$3 together with Clebsch-Gordan identities.

The main convergence statement is

$G = \mathrm{Spin}(d)$4

established by combining the asymptotic product and commutator properties of Toeplitz operators with the Kähler identities $G = \mathrm{Spin}(d)$5 and $G = \mathrm{Spin}(d)$6. A parallel construction gives square-matrix regularizations $G = \mathrm{Spin}(d)$7 acting on square Toeplitz operators $G = \mathrm{Spin}(d)$8 built from elements of $G = \mathrm{Spin}(d)$9, with ordinary commutators replacing the rectangular bracket.

Commutators and curvature

The appendix computes the commutator of two regularized covariant derivatives in the large-$\mathrm{Spin}^c(d)$0 limit:

$\mathrm{Spin}^c(d)$1

with $\mathrm{Spin}^c(d)$2 and $\mathrm{Spin}^c(d)$3 the curvature of the bundle associated to representation $\mathrm{Spin}^c(d)$4. All other contributions — including those from the first-order Poisson-bracket term and the representation-independent part $\mathrm{Spin}^c(d)$5 of the star-product coefficient — cancel or vanish as $\mathrm{Spin}^c(d)$6. The result is consistent with the continuum relation $\mathrm{Spin}^c(d)$7, since the surviving term is proportional to the curvature of the representation bundle. Notably, the proof requires a nontrivial group-theoretic identity among products of Clebsch-Gordan coefficients, established via a completeness relation on $\mathrm{Spin}^c(d)$8.

Examples: $\mathrm{Spin}^c(d)$9 and $\Lambda = -(d-2)m^2$0

For the flat torus $\Lambda = -(d-2)m^2$1, the Dirac zero modes factorize into theta-function-like solutions on each $\Lambda = -(d-2)m^2$2 factor, labeled by $\Lambda = -(d-2)m^2$3 where $\Lambda = -(d-2)m^2$4 is the flux quantization integer (the quantizability condition requires all area ratios $\Lambda = -(d-2)m^2$5 to be rational). The embedding Toeplitz operators reduce to clock and shift matrices satisfying the 't Hooft-Weyl algebra $\Lambda = -(d-2)m^2$6. Because $\Lambda = -(d-2)m^2$7 on the flat torus, the general commutator formula implies $\Lambda = -(d-2)m^2$8, correctly reproducing $\Lambda = -(d-2)m^2$9.

For the unit sphere $U(N)$0 with the Wu-Yang monopole connection ($U(N)$1), the zero modes are holomorphic sections with dimension $U(N)$2 in charge-$U(N)$3 sector, and the Toeplitz operators $U(N)$4, $U(N)$5 form the $U(N)$6 algebra with $U(N)$7. Using $U(N)$8, the commutator evaluates to

$U(N)$9

i.e., proportional to the Spin(2) charge, exactly as expected from the continuum relation $$\nabla_{(a)} = R^{\langle v\rangle}_{(a)b}(g^{-1})\,\nabla_b$$0. Defining $$\nabla_{(a)} = R^{\langle v\rangle}_{(a)b}(g^{-1})\,\nabla_b$$1, the triple $$\nabla_{(a)} = R^{\langle v\rangle}_{(a)b}(g^{-1})\,\nabla_b$$2 closes on the $$\nabla_{(a)} = R^{\langle v\rangle}_{(a)b}(g^{-1})\,\nabla_b$$3 Lie algebra in the large-$$\nabla_{(a)} = R^{\langle v\rangle}_{(a)b}(g^{-1})\,\nabla_b$$4 limit, matching the continuum algebra of the covariant derivative interpretation. The $$\nabla_{(a)} = R^{\langle v\rangle}_{(a)b}(g^{-1})\,\nabla_b$$5 construction is also shown to reproduce the earlier surface results of Hattori, Mizuno, and Tsuchiya up to an overall sign convention.

Limitations and open questions

Several restrictions qualify the results. The construction applies only to closed connected Kähler manifolds admitting a quantizable symplectic form; Lorentzian signatures and generic pseudo-Riemannian geometries lie outside its scope. The regularization depends on a choice of isometric embedding $$\nabla_{(a)} = R^{\langle v\rangle}_{(a)b}(g^{-1})\,\nabla_b$$6, and different embeddings plausibly correspond to different regularizations; the authors state explicitly that verifying independence of physical quantities from this choice remains an open problem. The double limit $$\nabla_{(a)} = R^{\langle v\rangle}_{(a)b}(g^{-1})\,\nabla_b$$7, $$\nabla_{(a)} = R^{\langle v\rangle}_{(a)b}(g^{-1})\,\nabla_b$$8 with $$\nabla_{(a)} = R^{\langle v\rangle}_{(a)b}(g^{-1})\,\nabla_b$$9 is taken formally, and no quantitative estimate of finite-$\Gamma(E_{\mathrm{reg}})$0 errors beyond the $\Gamma(E_{\mathrm{reg}})$1 scaling is provided. Finally, the framework has not yet been applied to actual simulation data; computing quantum effects such as the effective action within this regularization, and ultimately extracting curved geometry from numerical configurations of the type IIB matrix model, remain to be carried out.

Conclusion

This work supplies the missing technical ingredient for quantitative studies of curved spacetime in the covariant derivative interpretation of the type IIB matrix model: a Hermitian, finite-size matrix representative $\Gamma(E_{\mathrm{reg}})$2 of the covariant derivative that converges to $\Gamma(E_{\mathrm{reg}})$3 at rate $\Gamma(E_{\mathrm{reg}})$4 and whose commutator reproduces the curvature algebra in the large-matrix limit. Verified explicitly on $\Gamma(E_{\mathrm{reg}})$5 (commutativity restored) and $\Gamma(E_{\mathrm{reg}})$6 ($\Gamma(E_{\mathrm{reg}})$7 closure), the construction extends the Berezin-Toeplitz approach from Riemann surfaces to arbitrary quantizable Kähler manifolds of dimension $\Gamma(E_{\mathrm{reg}})$8, opening the way to loop calculations and to geometric interpretations of Monte Carlo results in the matrix model.