🔥The 9th Round of Easy Loan, Earn $40 Reward is in progress❗️ ⏰ Promotion Period: January 15th - Feburary 15th, 2025 👉 Register now and check more details at gate.io/campaigns/358

At chemical accuracy, #QML based on Danish Khan's ultra-compact local many body distribution functionals (MBDF) reaches #QM9 sampling rates of nearly 40 molecules/core second. #QMugs (Kenneth Atz) performance is also promising! With Heini 💉💉💉

Adaptive hybrid density functionals cure the spin gap problem in DFT! Optimal HF exchange admixture ratios for specific molecules of arbitrary size and composition, obtained on the fly from amon based #QML models trained across #chemicalspace, result in much improved hybrid DFT

Adaptive atomic basis sets! New preprint live arxiv.org/abs/2404.16942 Amazing work by Danish Khan and Maximilianach! #MachineLearning models of optimal scaling factors render small basis sets more accurate - at negligible overhead. #COMPCHEM #Quantumchemistry #chemicalspace

It seems we're learning that deep learning is mostly about the data. If you want to know where it will really take off look to areas where you can continuously generate increasingly diverse but consistently high quality data. That leads you to quantum chemistry:

A busy and inspiring week of #chemical #compound space exploration is over! Big thanks to co-organisers, student helpers, speakers, poster presenters, participants, sponsors... and see you in 2 years!

Celebrating the acceptance of our paper introducing machine learning of adaptive hybrid density functionals in Science Advances ! 🥳🥳🥳 We demonstrated that the combination of ideas from functional data analysis with one of the most common approximations in density functional

With less than 1M compounds, our VQM24 dataset sets a new SOTA in terms of covering more comprehensively the compositional and configurational space of small molecules. Wonderful joint efforts by current and past team-members and collaborators. Hopefully useful for #ML &

Dive into the latest from our lab with our newly published article in J Chem Phys! 🌟 @SimonLeonKrug1 and I have been exploring new frontiers in atomistic simulations in terms of universal Alchemical Harmonic Approximation (AHA) based potentials which seamlessly transmutate

Leonard Susskind: "The second law of thermodynamics and statistical mechanics are as deep as the Higgs boson or particle physics." *Whispering*: "Actually they're deeper." "Hope I am not disowned by my particle physics friends." youtube.com/watch?v=D1RzvX…

Refining hybrid density functionals with machine learning to enhance electronic structure predictions Density functional theory has long served as the cornerstone of computational chemistry, offering a balance between accuracy and efficiency in describing molecular and material

Super pumped about the prospect of #MachineLearning the adaptive exact exchange admixture in hybrid #DFT approximations: It can even cure the infamous spin-gap problem (see below). Brilliant work out in Science Advances Ilya Kuprov by Danish Khan Alastair Price Bing Huang and Max Ach!

The second part of the ☕️TEA Challenge is now published in Chemical Science! link: doi.org/10.1039/D4SC06…

The first part of the ☕️TEA Challenge, featuring Model Analysis, has now also been published in Chemical Science! This concludes the Crash Testing Machine Learning Force Fields TEA Challenge. Congratulations to everyone involved! link: doi.org/10.1039/D4SC06…

We are already on Day 3 of the workshop "Density Functional Theory and Artificial Intelligence learning from each other" in sunny CECAM-HQ. This afternoon, Fang Liu (Emory University) & Michael Herbst (EPFL) will present their talks.

A new non-empirical double hybrid density functional which performs well for general thermochemistry, kinetics and non-covalent tasks while remaining stable for bond dissociations. Please share your thoughts! arxiv.org/abs/2503.22661

#NewPaper Introducing MuAPBEK, a new kinetic energy density functional for quantum chemistry. MuAPBEK has several desirable features. 🐯It is fully analytical: the functional has a closed-form expression and can be optimized for any electronic system via automatic