Abstract:Finding Ricci-flat (Calabi-Yau) metrics is a long standing problem in geometry with deep implications for string theory and phenomenology. A new attack on this problem uses neural networks to engineer approximations to the Calabi-Yau metric within a given Kähler class. In this paper we investigate numerical Ricci-flat metrics over smooth and singular K3 surfaces and Calabi-Yau threefolds. Using these Ricci-flat metric approximations for the Cefalú family of quartic twofolds and the Dwork family of quintic threefolds, we study characteristic forms on these geometries. We observe that the numerical stability of the numerically computed topological characteristic is heavily influenced by the choice of the neural network model, in particular, we briefly discuss a different neural network model, namely Spectral networks, which correctly approximate the topological characteristic of a Calabi-Yau. Using persistent homology, we show that high curvature regions of the manifolds form clusters near the singular points. For our neural network approximations, we observe a Bogomolov--Yau type inequality $3c_2 \geq c_1^2$ and observe an identity when our geometries have isolated $A_1$ type singularities. We sketch a proof that $\chi(X~\smallsetminus~\mathrm{Sing}\,{X}) + 2~|\mathrm{Sing}\,{X}| = 24$ also holds for our numerical approximations.

What problem does this paper attempt to address?

### Problems the paper attempts to solve The paper aims to solve the problem of finding Ricci - flat (Calabi–Yau) metrics. Specifically, the researchers use neural network techniques to approximate Calabi–Yau metrics in specific Kähler classes and perform numerical calculations on smooth and singular K3 surfaces and Calabi–Yau three - manifolds. Through these approximated Ricci - flat metrics, the researchers study geometric characteristic forms and explore the influence of different neural network models on the numerical stability of topological characteristics. ### Overview of the main content 1. **Background and motivation**: - Ricci - flat metrics satisfy the Einstein equation and describe empty spacetime without matter sources, so they are of great significance in physics. - Calabi–Yau spaces describe the extra spatial dimensions in string theory compactification. - Although many Calabi–Yau properties can be obtained by topological methods, specific Ricci - flat metrics are crucial for certain explicit calculations, such as α' corrections, determination of Kähler potential, calculation of Yukawa couplings, and inference of supersymmetry breaking in low - energy effective field theory. 2. **Research methods**: - Use neural network techniques to approximate Calabi–Yau metrics. - The research objects include quartic hypersurfaces in the Cefalú family and quintic three - manifolds in the Dwork family. - Use the persistent homology method to analyze the distribution of high - curvature regions. 3. **Main results**: - Through the metrics approximated by neural networks, the researchers observe that the topological characteristics calculated numerically are greatly influenced by the selected neural network model. - High - curvature regions form clusters near the singular points of the manifold. - The Bogomolov–Yau - type inequality $3c_2 \geq c_1^2$ is observed, and an equality relationship is found when the geometry has isolated A1 - type singularities. - It is proved that $\chi(X\setminus \text{Sing}X)+ 2|\text{Sing}X| = 24$ holds under numerical approximation. ### Formula display - **Kähler metric**: \[ g_{a\bar{b}}=\partial_a\partial_{\bar{b}}K(z,\bar{z}) \] - **Ricci tensor**: \[ R_{a\bar{b}} = -\partial_a\partial_{\bar{b}}\log\det g \] - **Euler density**: \[ c_2(J_{\text{CY}})=\frac{1}{8\pi^2}\sqrt{g}R_{a\bar{b}c\bar{d}}R^{a\bar{b}c\bar{d}}d^4z \] - **Bogomolov–Yau inequality**: \[ 3c_2(J)-c_1(J)^2\geq0 \] - **Euler number calculation**: \[ \chi(X)=\deg c_{\text{SM}}(X) \] - **Fulton class and Milnor number relationship**: \[ (-1)^{\dim X}\deg(c_F(X)-c_{\text{SM}}(X))=\sum_{p\in\text{Sing}X}\mu_X(p) \] ### Conclusion Through neural network techniques, the researchers successfully approximated Ricci - flat metrics on smooth and singular K3 surfaces and Calabi–Yau three - manifolds, and verified the effectiveness of these metrics in calculating topological characteristics and curvature distributions. These results provide new tools and perspectives for further research on Calabi–Yau metrics and their applications in string theory.

Machine Learned Calabi-Yau Metrics and Curvature

Bach-flat Critical Metrics for Quadratic Curvature Functionals

Levi-Civita Ricci-flat Metrics on Compact Complex Manifolds

Lectures on Numerical and Machine Learning Methods for Approximating Ricci-flat Calabi-Yau Metrics

Calabi-Yau metrics through Grassmannian learning and Donaldson's algorithm

New results on the geometry of the moduli space of Riemann surfaces

Curvatures of Moduli Space of Curves and Applications

Learning Group Invariant Calabi-Yau Metrics by Fundamental Domain Projections

Not So Flat Metrics

Calabi-Yau Manifolds with Isolated Conical Singularities

Calabi Flow, Geodesic Rays, and Uniqueness of Constant Scalar Curvature Kähler Metrics

Machine learning Sasakian and $G_2$ topology on contact Calabi-Yau $7$-manifolds

cymyc -- Calabi-Yau Metrics, Yukawas, and Curvature

Calabi-Yau metrics on rank two symmetric spaces with horospherical tangent cone at infinity

Warped quasi-asymptotically conical Calabi-Yau metrics

Harmonic $1$-forms on real loci of Calabi-Yau manifolds

Calabi-Yau Four/Five/Six-folds as $\mathbb{P}^n_\textbf{w}$ Hypersurfaces: Machine Learning, Approximation, and Generation

Calabi-Yau metrics on Kummer surfaces as a model glueing problem

Distinguishing Calabi-Yau Topology using Machine Learning

Machine learning Calabi-Yau four-folds

Complete Kähler manifolds with nonnegative Ricci curvature