Fascinating paper, showing that transformers are energy-based models in disguise .. And this insight leads to an efficient decoding algorithm

Checkout Generative Hierarchical Materials Search (GenMS) – a framework for generating crystal structures from natural language. Website: generative-materials.github.io Paper: arxiv.org/abs/2409.06762

😍 Our work on the “Chemical-motif characterization of short-range order with E(3)-equivariant graph neural networks” has been published in Nature - npj Computational Materials npj Journals 😍 Check out the findings in our paper here: nature.com/articles/s4152…

(1/3) Sharing preprint by Shengli (Bruce) Jiang | 蒋晟立 on combining physical baseline models (Gaussian chain theory) and ML to boost transferability of property predictions for architecturally+compositionally complex polymers. #MachineLearning #compchem #AI4science shorturl.at/jUOhA

ELLIS Machine Learning for Molecules workshop December 6, 2024, HYBRID Paper submission deadline: November 1st. Program chairs: Francesca Grisoni @[email protected] an me Webpage: moleculediscovery.github.io/workshop2024/ We hope to see you all there!

📢Join us for a new Quantum Spain #Seminar! 🚀 Title: Quantum Convolutional Neural Networks are (Effectively) Classically Simulable 🗣️Speaker: Pablo Bermejo, PhD Student at DIPC (Spain) and Los Alamos (USA) 🗓️Date: Sept 24, 2024 🔗 Register here: events.teams.microsoft.com/event/06fd2ebc…

Harnessing machine learning to understand how electric fields affect materials Joll, Schienbein, Rosso, and Blumberger have developed a new method to model how condensed phase systems respond to electric fields. This is crucial for many applications, from energy storage and

Advancing molecular property prediction with fractional denoising Deep learning holds great promise for speeding up molecular screening in drug discovery and material design, but limited labeled data remains a challenge. A new study by Yuyan Ni and coauthors introduces

MolPipeline: Integration of RDKit and Scikit-learn for enhanced molecular machine learning workflows While major discoveries often capture the spotlight, contributions like MolPipeline play also a vital role in advancing scientific progress. In their recent work, Jochen Sieg and

Accelerating functional coatings innovation with data-driven approaches Data-driven methods are playing an increasingly significant role in the development of next-generation materials, particularly in functional coatings. A recent paper by Kai Xu and coauthors reviews how

Looking forward to this new seminar from the ELLIS unit Cambridge by Alex Tong. Next Tuesday at 3pm UK time. Zoom details in the link. Add it to your calendar!

MOCCA: A Python package for automated chromatogram processing In chemical reaction analysis, efficient and accurate data processing tools are essential to advancing research. The MOCCA Python package, developed by Jan Obořil and colleagues, offers an automated solution for

#compchem Good read: Catalysis in the digital age: Unlocking the power of data with machine learning doi.org/10.1002/wcms.1…

Crystalyze: Generative machine learning for materials crystal structure determination Machine learning is increasingly being applied to the field of crystal structure determination from powder X-ray diffraction (PXRD) patterns, a widely used technique in materials

Fantastic piece of quantum computing engineering 👏🏻

PerQueue: Streamlining workflows in computational materials science PerQueue is a Python-based workflow manager developed to help researchers in computational materials science manage complex workflows more efficiently. Created by Benjamin Heckscher Sjølin and coauthors,

Boosting dye-sensitized solar cell efficiency through machine learning-guided optimization Improving the efficiency of solar cells is a key challenge in the push for more sustainable energy solutions. Liao et al. have applied machine learning to optimize dye-sensitized solar

Collective variables without descriptors? In the latest preprint by Jintu Zhang, Luigi Bonati and Enrico Trizio we used graph neural networks to design CVs for enhanced sampling directly as a function of atomic positions 🧵⤵️ arxiv.org/abs/2409.07339 #compchem #mlcolvar #plumed

Professor Yet-Ming Chiang has been named to the Forbes Sustainability Leaders list, the magazine’s inaugural tally of 50 leaders demonstrating exceptional ambition, innovation, and recent, tangible impact that is both scalable and sustainable. buff.ly/3XBtoj7

Excited to share our new tutorial on using diffusion models to enhance the resolution of microscopy images!🧬🔬 It covers the key fundamentals with a full, step-by-step code implementation of diffusion models from scratch. Take a look! arxiv.org/abs/2409.16488 Giovanni Volpe