Update README

README.md

@@ -1,3 +1,31 @@
 ---
 license: mit
 ---
+
+# SimPoly
+
+SimPoly is a fast and scalable machine learning force field (MLFF) for ab initio prediction of polymer properties. This repository contains the trained model weights and benchmark datasets from our paper. Refer to the accompanying GitHub repository for instructions and usage examples.
+
+**Key Features:**
+- Accurately predicts polymer densities and glass transition temperatures without experimental fitting
+- Outperforms classical force fields for a broad range of polymers
+- Includes benchmark data: experimental bulk properties for 130 polymers and quantum-chemical training datasets
+
+**Resources:**
+- [GitHub Code](https://github.com/microsoft/simpoly)
+- [Paper on arXiv](https://arxiv.org/abs/2510.13696)
+
+**Citation**
+
+If you use this work, please cite:
+```
+@misc{simm2025simpolysimulationpolymersmachine,
+      title={SimPoly: Simulation of Polymers with Machine Learning Force Fields Derived from First Principles}, 
+      author={Gregor N. C. Simm and Jean Hélie and Hannes Schulz and Yicheng Chen and Guillem Simeon and Anna Kuzina and Ernesto Martinez-Baez and Piero Gasparotto and Gabriele Tocci and Chi Chen and Yatao Li and Lixue Cheng and Zun Wang and Bichlien H. Nguyen and Jake A. Smith and Lixin Sun},
+      year={2025},
+      eprint={2510.13696},
+      archivePrefix={arXiv},
+      primaryClass={physics.chem-ph},
+      url={https://arxiv.org/abs/2510.13696}, 
+}
+```