Tuning moiré lattice geometry and topology in twisted transition metal dichalcogenides using machine-learning assisted ab initio calculations

Twisted transition metal dichalcogenides (TMDs) have emerged as a promising platform for exploring electronic topology, strong correlations, and optically excited states. In this work, we combine ab initio calculations with machine learning and data science techniques to reveal how twisted double bilayer moiré TMDs exhibit tunable lattice geometries and band topologies as a function of twist angle and external fields. This tunability provides a rich landscape for investigating competing correlated phases and discovering a variety of topological states beyond the 1+1 moiré limit. Our approach establishes a framework for the efficient exploration of other moiré materials through machine-learning-assisted ab initio calculations, offering connections to recent experimental observations and expanding the possibilities for designing next-generation quantum materials.