Machine Learning Interatomic Potential– tag –
-
Matlantis and NVIDIA ALCHEMI Cut ENEOS Catalyst Discovery from Years to Months, Screening 100 Million Structures
AiThority Unknown Overview Matlantis announced that the combination of NVIDIA ALCHEMI and its Matlantis PFP (universal machine learning interatomic potential) significantly accelerated catalyst materials discovery for ENEOS Holdings. Thi... -
Machine Learning Potentials Grapple with Long-Range Interactions in Atmospheric Modeling: A DTU Deep Dive
DTU Research Database Denmark Overview Researchers at the Technical University of Denmark (DTU) evaluated AIMNet2 and PaiNN, two machine-learned interatomic potentials (MLIPs), for modeling molecular collisions critical to atmospheric cl... -
Hessian-Based Molecular Conformation Augmentation Improves Scalability and Efficiency of Machine Learning Interatomic Potentials
arXiv (Preprint Server) International Overview This preprint proposes two Hessian-derived data augmentation schemes, isotropic Gaussian displacement (UniAug) and normal mode-weighted displacement (ModeAug), for machine-learning interatom... -
Physics-Aware AI Indispensable for Materials Research: Performance Gaps Highlighted in Thermal Conductivity Prediction
Columbia Engineering USA Overview Columbia Engineering emphasizes the critical need for physics-aware AI in materials science, particularly for machine learning interatomic potentials (MLPs) that predict macroscopic properties from quant... -
Equivariant Graph Neural Network Interatomic Potentials Accelerate Nanomaterials Simulation by 100x
nano-matter.com International Overview Equivariant Graph Neural Network (GNN) interatomic potentials are accelerating nanomaterials research by predicting atomic forces and energies with DFT-comparable accuracy, while reducing computatio... -
GRACE-OFF Achieves High-Precision Machine-Learned Interatomic Potentials for Organic Liquids via GRACE Architecture
Journal of Chemical Theory and Computation | ACS Publications USA Overview This study introduces GRACE-OFF, a machine-learned interatomic potential (MLIP) built on the Graph Atomic Cluster Expansion (GRACE) neural network architecture, d... -
LAMMPS Integrates DeePMD-kit, NequIP for Quantum-Accurate ML Potentials, Enhancing Molecular Dynamics Simulations
LAMMPS Molecular Dynamics Simulator USA Overview LAMMPS, a widely used open-source molecular dynamics simulator, now supports cross-code frameworks and interoperability tools for machine learning interatomic potentials (MLPs), including ... -
AIP Publishing Reveals New Vibrational Entropy Metric Dramatically Improves Empirical Interatomic Potential Accuracy by 80%, Revolutionizing MLIP Validation
AIP Publishing USA Overview A new study published by AIP Publishing introduces a novel method for developing and validating empirical interatomic potentials, achieving an 80% error reduction in dynamic properties compared to existing MOF... -
AI-Driven Hyperdynamics Method Revolutionizes Atomic Event Simulation for Material Durability and Aging Research
Nature Communications Unknown Overview A novel machine learning-based hyperdynamics method dramatically accelerates atomic motion simulations, transforming computational materials discovery. This automated approach for developing interat... -
GitHub’s AI4Science Repository Curates Essential AI Tools, Libraries, and Datasets to Accelerate Scientific Discovery, Including Materials Science
GitHub - ai4s-research/awesome-ai-for-science USA Overview The 'ai4s-research/awesome-ai-for-science' GitHub repository now provides a curated list of AI tools, libraries, papers, datasets, and frameworks designed to accelerate scientifi...
