Get Started
Home
Topics
Search
Library
Research questionHow can machine-learning interatomic potentials support Hessian-dependent applications using only energy and force data?Standard potentials are commonly trained on energies and atomic forces, while Hessian information is needed for several downstream analyses. Incorporating Hessians directly can require architectural changes, higher-order differentiation, and substantial computational and memory resources.
AI
Machine Learning
Research Paper
Technology
Latest papersRecent research connected to this question, newest first.Hessian-based molecular conformation augmentation for a scalable and efficient strategy of machine learning interatomic potentialsThe source addresses ML interatomic potentials and evaluates Hessian-derived data augmentation on equilibrium and non-equilibrium datasets. Its evidence concerns improvements in model accuracy and task-specific guidance; it does not establish performance for every potential architecture or downstream application.research paper · Sep 4, 2026
Related questions
How can machine-learning interatomic potentials capture long-range electrostatics and polarization in ionic, polar, and interfacial systems?Under what genericity, overlap, and connectivity conditions can sparse cutoff-based multi-layer message passing universally approximate interatomic potential energy surfaces?How can physics-informed neural networks handle stiff, multiscale ion-electronic transport while supporting inverse estimation?How can AI capture polymers’ stochastic, multiscale structure for reliable property prediction and molecular design?