A New Type of Saddle in the Euclidean IKKT Matrix Model and Its Emergent Geometry

Henry Liao, Reishi Maeta · 2025

Read A New Type of Saddle in the Euclidean IKKT Matrix Model and Its Emergent Geometry by Henry Liao, Reishi Maeta (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2512-03161.txt.

Introduction

A matrix model of physics may hold new shapes of space. Henry Liao and Reishi Maeta study the Euclidean IKKT matrix model in this 2025 work. They look at its equation of motion and seek new steady solutions, called saddles.

They find a new type of saddle that exists only as the matrix size grows without bound. Assuming the matrices match a known symmetry algebra, they identify a unique solution. From it, a geometry emerges. Such work probes how space itself might arise from matrices in quantum gravity research. Emergent geometry from matrices remains a bold idea, and each exact solution tests its reach.

Catalog record

CREATORHenry Liao, Reishi Maeta
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE22 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2512-03161

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

A New Type of Saddle in the Euclidean IKKT Matrix Model and Its Emergent Geometry Henry Liao, Reishi Maeta Abstract We study the equation of motion of the Euclidean IKKT matrix model, and realize a new type of classical saddle that only exists in N → ∞ N\rightarrow\infty limit. Under the assumption that the matrices are the generators of 𝔰 ​ 𝔬 ​ ( n , m ) \mathfrak{so}(n,m) , we identify a unique solution, that is, 𝔰 ​ 𝔬 ​ ( 1 , 3 ) \mathfrak{so}(1,3) . Even though it has 6 6 generators and thus 6 6 non-zero matrices, they are not independent due to the 2 2 Casimir constraints in 𝔰 ​ 𝔬 ​ ( 1 , 3 ) \mathfrak{so}(1,3) . Exploiting the Lie-algebraic structure and the Casimir constraints, we derive a four-dimensional space that a test scalar propagates on. The associated metric possesses SU ⁡ ( 2 ) \mathrm{SU}(2) isometry, which is closely related to the Taub–NUT/Bolt geometry and, more broadly, to black hole physics. I Introduction As a promising proposal toward non-perturbative formulation of type IIB superstring theory, the IKKT matrix model 22 ; 3 is expected to describe non-perturbative objects such as a D3 -brane that we live on. In fact, this zero-dimensional matrix model suggests a picture that four-dimensional spacetime can be emerged dynamically 24 ; 2 ; 33 ; 9 . In addition, its low energy effective theory is closely related to general relativity and can be unitary 21 . There are other attempts to understand the IKKT matrix model. A holographic dictionary between type IIB supergravity and the IKKT matrix model has been established 14 ; 13 . For the so-called polarised IKKT matrix model 7 , in which special deformations are added while preserving maximal supersymmetry, there are similar discussions 20 ; 19 ; 27 ; 28 . For other types of deformations, various studies have shown the emergence of four-dimensional near-realistic spacetime, which has FLRW metric 38 ; 37 ; 5 or even chiral fermions 32 ; 10 . Among these various developments, from the viewpoint of describing the real world in which we live, the original IKKT matrix model without mass deformations is the most suitable for detailed analysis. This is because the IKKT matrix model is derived either (1) by the large- N reduction of ten-dimensional \mathcal{N}=1 supersymmetric Yang–Mills theory 17 ; 35 ; 18 , or (2) by matrix regularization of the Green–Schwarz action of type IIB superstring theory 22 (for type IIA superstring theory, see also 16 ; 4 ), and in both derivations no mass term is present. However, for the massless theory, only a limited number of classical solutions are currently known 22 ; 30 ; 11 ; 36 , let alone solutions describing four-dimensional spacetime. Motivated by this situation, we try to find a new saddle of the original IKKT matrix model which is defined in Euclidean signature. More concretely, we examined whether configurations of the form A_{\mu}=M_{ab} , constructed from the generators M_{ab} of \mathfrak{so}(n,m) , satisfy the equations of motion ( 2 ). As a result, we found that among these cases, \mathfrak{so}(1,3) is the only solution. Furthermore, we analysed which geometry is emergent from such a saddle. It turns out that the resulting geometry describes a four-dimensional Euclidean spacetime with \mathrm{SU}(2) isometry, which bears a strong resemblance to the Taub–Nut/Bolt geometry 40 ; 31 . So far no example in which black hole physics emerges from the IKKT matrix model has been known, and the \mathfrak{so}(1,3) solution may constitute the first such example. In the beginning of Sec. II , we briefly mention our setup, and show the action and the equation of motion in the IKKT matrix model. Then, we mention a standard statement related to the equation of motion in the Euclidean IKKT matrix model, see Sec. II.1 . In there, we point out a loophole to the statement occurs when the matrices are infinite-dimensional. This motivates us to consider, in Sec. II.3 , an ansatz that the matrices are infinite dimensional. More specifically, they are generators of \mathfrak{so}(n,m) with n,m\geq 1 , which only have infinite dimensional unitary representations. 1 1 1 The \mathfrak{so}(n,m) is different from the spacetime Lorentz symmetry in general. We find that, under our ansatz, \mathfrak{so}(1,3) is the only possibility. This is a complete new class of solution, see II.2 . To analyse its emergent geometry, we consider a test scalar matrix to find the effective metric in semi-classical limit. This procedure aligns with that in 39 , and we provide the detail in Sec. III . Then, we conclude in IV . II Classical saddles of Euclidean Bosonic IKKT matrix model In the Euclidean IKKT matrix model, the action of the bosonic sector reads \displaystyle S_{\text{IKKT,B}}=-\frac{1}{4}\mathrm{tr}\left(\delta_{\mu\rho}\delta_{\nu\sigma}\left[A^{\mu},A^{\nu}\right]\left[A^{\rho},A^{\sigma}\right]\right). (1) The equation of motion for A^{\mu} is given by \displaystyle\delta_{\mu\rho}\left[A^{\mu},\left[A^{\rho},A^{\sigma}\right]\right]=0, (2) which can also be regarded as equation of motion of the full IKKT matrix model but with the assumption that the fermionic matrices set to 0 . Note that during the derivation, we have assumed that the cyclic property of trace is always valid, and this is assumed throughout this paper. II.1 A Standard Statement and Its Loophole A standard statement of the solutions for the Euclidean equation of motion ( 2 ) goes as follows. Starting with ( 2 ), we multiply from left an \delta_{\nu\sigma}A^{\nu} and take trace which yields \begin{split}0&=\mathrm{tr}\left(\delta_{\nu\sigma}A^{\nu}\delta_{\mu\rho}\left[A^{\mu},\left[A^{\rho},A^{\sigma}\right]\right]\right)\\ &=-\mathrm{tr}\left(\delta_{\mu\rho}\delta_{\nu\sigma}\left[A^{\mu},A^{\nu}\right]\left[A^{\rho},A^{\sigma}\right]\right)\\ &=\mathrm{tr}\left(\left(i\left[A_{\mu},A_{\nu}\right]\right)\left(i\left[A_{\mu},A_{\nu}\right]\right)\right).\end{split} (3) Since A_{\mu} ’s are Hermitian matrices, the combination i\left[A_{\mu},A_{\nu}\right] is also Hermitian; for each pair of \mu,\nu , i\left[A_{\mu},A_{\nu}\right] has real eigenvalues and thus \mathrm{tr}\left(i\left[A_{\mu},A_{\nu}\right]\right)^{2} is non-negative. Hence, the last line of ( 3 ) can only vanish if all eigenvalues of i\left[A_{\mu},A_{\nu}\right] vanish for all \mu,\nu , meaning A^{\mu} ’s commute. Even though this statement sounds solid, it implicitly assumes A^{\mu} ’s are finite in size. In fact, it is clear that Heisenberg algebra, which is defined by \left[X,Y\right]=Z , \left[X,Z\right]=0 and \left[Y,Z\right]=0 , is a solution of ( 2 ). Moreover, some nilpotent and solvable Lie algebras are solutions as well 11 . The loophole stems from that the argument intrinsically assumes that the trace operation is only over finite terms, that is, having finite unitary representations. However, the exceptional cases mentioned above do not have finite dimensional unitary representations, which protects them from the argument. Hence, to look for other types of classical saddles, one must consider cases where A^{\mu} ’s are infinite dimensional. II.2 Classification of the IKKT saddles Before discussing our solution, we make a brief detour on a classification of the saddles that can occur in ( 2 ). With simple observation, ( 2 ) contains three operations: (a) the inner commutator, (b) the outer commutator and (c) the summation over spacetime indices. To satisfy the equation of motion ( 2 ), we can have the solution to be vanished at either one of the steps, meaning three different categories of solutions. The solutions mentioned previously fall either into (a) or (b). In particular, the case where \left[X,Y\right]=0 falls into (a), while the cases of Heisenberg, solvable and nilpotent Lie algebras are contained in (b). There is no known third class solution so far, until now. In particular, the solution we

Read the complete text