Deep multi-task mining Calabi-Yau four-folds

Abstract

We continue earlier efforts in computing the dimensions of tangent space cohomologies of Calabi-Yau manifolds using deep learning. In this paper, we consider the dataset of all Calabi-Yau four-folds constructed as complete intersections in products of projective spaces. Employing neural networks inspired by state-of-the-art computer vision architectures, we improve earlier benchmarks and demonstrate that all four non-trivial Hodge numbers can be learned at the same time using a multi-task architecture. With 30% (80%) training ratio, we reach an accuracy of 100% for $h^{(1,1)}$ and 97% for $h^{(2,1)}$ (100% for both), 81% (96%) for $h^{(3,1)}$, and 49% (83%) for $h^{(2,2)}$. Assuming that the Euler number is known, as it is easy to compute, and taking into account the linear constraint arising from index computations, we get 100% total accuracy.

15 pages; additional details, references updated

References in corpus (14)

• Array Programming with NumPy
• Language Models are Few-Shot Learners
• Computational complexity of the landscape I
• Explaining Neural Scaling Laws
• Evolving neural networks with genetic algorithms to study the String Landscape
• Branes with Brains: Exploring String Vacua with Deep Reinforcement Learning
• Fibrations in CICY Threefolds
• Topological Invariants and Fibration Structure of Complete Intersection Calabi-Yau Four-Folds
• Getting CICY High
• Machine Learning Calabi-Yau Four-folds
• Machine learning for complete intersection Calabi-Yau manifolds: a methodological study
• Explore and Exploit with Heterotic Line Bundle Models
• Inception Neural Network for Complete Intersection Calabi-Yau 3-folds
• Deep Miner: A Deep and Multi-branch Network which Mines Rich and Diverse Features for Person Re-identification

Cited by in corpus (1)

• Characterizing 4-string contact interaction using machine learning