PaddleMaterials is a data-mechanism dual-driven, foundation model development and deployment, end to end toolkit based on PaddlePaddle deep learning framework for materials science.

[ICLR 2026] Advancing Universal Deep Learning for Electronic-Structure Hamiltonian Prediction of Materials

This repository is a collection of my thoughts, summaries, and reflections on literature, websites, and tutorials that explore the intersection of materials science and artificial intelligence.