Sign in

Jan Hermann

@hrmnn.net
1.8K followers 149 following 152 posts

Computational chemistry & physics, electrons, deep learning 🚲☕️♟️ Microsoft Research AI for Science · hrmnn.net

PostsRepliesMedia
Jan Hermann @hrmnn.net · 26/08/2026
Trying to model bond dissociation accurately and running into a wall with both wavefunction and DFT methods? Orbformer, our pretrained wavefunction ansatz, makes it easier with deep QMC. Now looking into using it to generate bond dissociation training data for Skala. www.nature.com/articles/s41...
nature.com
An ab initio foundation model of wavefunctions that accurately describes chemical bond breaking - Nature Communications
Orbformer is a quantum Monte Carlo foundation model pretrained on thousands of molecular structures. It can be fine-tuned to yield accurate wavefunctions, including for multireference phenomena, with ...
010
Jan Hermann @hrmnn.net · 24/08/2026
Skala, our XC functional, now outperforms the best global hybrids on GMTKN55, at the cost of a semi-local functional. Coming soon to many popular DFT packages, starting with @cp2k.bsky.social, Psi4, ORCA by @faccts.de, VASP, and @fhi-aims.bsky.social
171
Jan Hermann @hrmnn.net · 22/04/2026
Today we're sharing a major Skala update—new paper and model release. Skala is a deep-learning exchange-correlation functional for DFT that reaches 2.8 kcal/mol on GMTKN55, wins 32 of its 55 subsets, and leads on accuracy short of double hybrids at semi-local DFT cost
1136
Jan Hermann @hrmnn.net · 01/12/2025
📢 Hiring into three new roles in the OneDFT team at MSR AI for Science! 💼 ⬇️ Join our mission to make DFT accurate and reliable, learn more at aka.ms/dft
careerhub.microsoft.com
Senior Research Engineer in DFT for Materials Science, AI for Science | Microsoft Careers
Implement and maintain evaluation pipelines for exchange correlation functionals for materials using software packages like VASP, CP2K, QuantumEspresso, FHI-aims, PySCF, or similar Work cross-function...
120
Jan Hermann @hrmnn.net · 09/10/2025
Our neural-network XC functional, Skala, is available in the cloud in Azure AI Foundry, on PyPI as an open-source Python package with hookups to PySCF and ASE, and via the C++ library GauXC for any third-party DFT code. If you find anything interesting about Skala, please let us know, we're curious!
github.com
GitHub - microsoft/skala: Skala exchange-correlation functional
Skala exchange-correlation functional. Contribute to microsoft/skala development by creating an account on GitHub.
051
Jan Hermann @hrmnn.net · 04/08/2025
Simulating molecules and materials accurately is one thing, knowing which molecules and materials to look at is another. Look at these new roles for the latter!
030
Jan Hermann @hrmnn.net · 03/08/2025
I benefited massively from www.ipam.ucla.edu/programs/lon.... I got into ML for science through that program. Now IPAM may be gone mathstodon.xyz/@tao/1149568...
mathstodon.xyz
Terence Tao (@tao@mathstodon.xyz)
The current administration in the US has, through various funding agencies such as the NSF and NIH, has recently suspended virtually all federal grants to my home university, UCLA (including my own p...
262
Jan Hermann @hrmnn.net · 08/07/2025
Interested in our mission to make DFT more accurate and push what’s possible in quantum chemistry? Do you want to directly contribute? We're hiring a senior software engineer and a senior researcher: jobs.careers.microsoft.com/global/en/jo... jobs.careers.microsoft.com/global/en/jo...
0187
Jan Hermann @hrmnn.net · 06/07/2025
@chrislhayes.bsky.social you achieved what I would have thought impossible. In just the first three chapters of your book you made my phone seem so disgusting that I’ve barely touched it in the last few days
010
Jan Hermann @hrmnn.net · 26/06/2025
🚀 Strong correlation is the Everest of quantum chemistry. Next to the coupled cluster highway, the multireference molecular terrain is underserved—gravel roads and promenades. With Orbformer, we're building a new infrastructure by marrying neural network wavefunctions with cost amortization at scale
1154
Jan Hermann @hrmnn.net · 18/06/2025
🚀 After two+ years of intense research, we’re thrilled to introduce Skala — a scalable deep learning density functional that hits chemical accuracy on atomization energies and matches hybrid-level accuracy on main group chemistry — all at the cost of semi-local DFT ⚛️🔥🧪🧬
37225
Reposted by Jan Hermann
Wessel @wessel.ai · 21/05/2025
Big news!! 🚀 Aurora, a foundation model for the Earth system, has been published in @nature.com. A massive congrats to the whole team! Very proud of this achievement, and thrilled to see that it’s finally out there. 😊 www.nature.com/articles/s41...
nature.com
A foundation model for the Earth system - Nature
Aurora, a new large-scale foundation model trained on more than one million hours of diverse geophysical data, outperforms operational forecasts in predicting air quality, ocean wave dynamics, tropica...
02611
Jan Hermann @hrmnn.net · 02/05/2025
7 years since finishing Phd, and some of the work still keeps coming out. @fhi-aims.bsky.social is a great piece of electronic structure software, and I'm proud I could have contributed a little bit arxiv.org/abs/2505.00125
arxiv.org
Roadmap on Advancements of the FHI-aims Software Package
Electronic-structure theory is the foundation of the description of materials including multiscale modeling of their properties and functions. Obviously, without sufficient accuracy at the base, relia...
2131
Reposted by Jan Hermann
Randall Munroe @xkcd.com · 25/04/2025
PhD Timeline xkcd.com/3081
5865979720444
Reposted by Jan Hermann
Nicholas Gao @n-gao.bsky.social · 09/04/2025
Crucially, our method is not limited to biochemical compounds but also extends to computationally challenging organometallic compounds with up to 180 electrons where we - like the previous results - match or surpass the gold-standard CCSD(T)/CBS.
131
Jan Hermann @hrmnn.net · 04/04/2025
Desperately searching for silver lining here. A huge global economic downturn might be good for the environment?
010
Reposted by Jan Hermann
dan sinker @dansinker.com · 03/04/2025
Guess where they got their weird trade deficit math from?
A tweet by "rohit" discusses the application of AI in geopolitics, specifically focusing on tariffs, and includes a table comparing tariff rates between China and the US.
25271112185
Reposted by Jan Hermann
Serge Belongie @serge.belongie.com · 30/03/2025
Would you present your next NeurIPS paper in Europe instead of traveling to San Diego (US) if this was an option? Søren Hauberg (DTU) and I would love to hear the answer through this poll: (1/6)
docs.google.com
NeurIPS participation in Europe
We seek to understand if there is interest in being able to attend NeurIPS in Europe, i.e. without travelling to San Diego, US. In the following, assume that it is possible to present accepted papers ...
6279161
Reposted by Jan Hermann
Frank Noe @franknoe.bsky.social · 26/03/2025
Simulating full quantum mechanical ground- and excited state surfaces with deep quantum Monte Carlo by Zeno Schätzle, Bernat Szabo and Alice Cuzzocrea. arxiv.org/abs/2503.19847 🧵⬇️
2316
Reposted by Jan Hermann
Jörg Enderlein @joergenderlein.bsky.social · 23/03/2025
I was invited, for July, to a scienctific meeting in Telluride (Colorado). I was told it is a wonderful place for a conference, and I was very much looking forward to an exciting meeting in a fantastic environment. I have just canceled my participation, despite having booked already everything.
theguardian.com
French scientist denied US entry after phone messages critical of Trump found
France’s research minister said the scientist was traveling to Houston for a conference when his phone was searched
46823
Reposted by Jan Hermann
Marwin Segler @marwinsegler.bsky.social · 11/02/2025
The transition-metal catalyst reduced the activation barrier to convert the cis into the trans isomer.
051
Reposted by Jan Hermann
Belinda Barnet @manjusrii.bsky.social · 21/01/2025
Donate to Wikipedia. Do it today, and every time one of the broligarchs has a tantrum because he can’t buy it
5138361908
Reposted by Jan Hermann
Daniel Zuegner @danielzuegner.bsky.social · 16/01/2025
Super excited to share that the MatterGen code is now public on GitHub! github.com/microsoft/ma...
github.com
GitHub - microsoft/mattergen: Official implementation of MatterGen -- a generative model for inorganic materials design across the periodic table that can be fine-tuned to steer the generation towards...
Official implementation of MatterGen -- a generative model for inorganic materials design across the periodic table that can be fine-tuned to steer the generation towards a wide range of property c...
01710
Reposted by Jan Hermann
Tian Xie @xie-tian.bsky.social · 16/01/2025
Excited to finally announce the publication of MatterGen on Nature. MatterGen represents a new paradigm of materials design with generative AI. We are releasing the code of MatterGen under MIT license. Look forward to seeing how the community will use the tool and build on top of it.
1129
Jan Hermann @hrmnn.net · 31/12/2024
45²
231
Jan Hermann @hrmnn.net · 18/12/2024
Super cool work
070
Jan Hermann @hrmnn.net · 15/12/2024
As for word limits in proposals—Rousseau’s en.wikipedia.org/wiki/Discour... apparently wasn’t considered in the 1754 essay contest it was written for because it went over the limit. The tradition is old
en.wikipedia.org
Discourse on Inequality - Wikipedia
140
Reposted by Jan Hermann
Frank Noe @franknoe.bsky.social · 06/12/2024
Super excited to preprint our work on developing a Biomolecular Emulator (BioEmu): Scalable emulation of protein equilibrium ensembles with generative deep learning from @msftresearch.bsky.social ch AI for Science. www.biorxiv.org/content/10.1...
21441147
Reposted by Jan Hermann
Claudio Zeni @claudiozeni.bsky.social · 03/12/2024
🚨Our Machine Learning Force Field Mattersim is now available! 🚨 Check it out here 👇 msft.it/6013oBZLt The force field is designed to be used on a vast range of temperatures and pressures, try it yourself :) Feedback and suggestions are very welcome!
msft.it
GitHub - microsoft/mattersim: MatterSim: A deep learning atomistic model across elements, temperatures and pressures.
MatterSim: A deep learning atomistic model across elements, temperatures and pressures. - microsoft/mattersim
58125
Reposted by Jan Hermann
Jay 🦋 @jay.bsky.team · 19/11/2024
Bluesky now has over 20M people!! 🎉 We've been adding over a million users per day for the last few days. To celebrate, here are 20 fun facts about Bluesky:
307013027816061
Reposted by Jan Hermann
Volker Blum @aimsduke.bsky.social · 15/11/2024
#chemsky #compchemsky The big day is here. I just received word that the final version of our 50+ coauthor "Roadmap on Methods and Softwarre for Electronic Structure Based Simulations in Chemistry and Materials" is out! Thank you to all involved! iopscience.iop.org/article/10.1...
iopscience.iop.org
Roadmap on methods and software for electronic structure based simulations in chemistry and materials - IOPscienceSearch
Roadmap on methods and software for electronic structure based simulations in chemistry and materials, Volker Blum, Ryoji Asahi, Jochen Autschbach, Christoph Bannwarth, Gustav Bihlmayer, Stefan Blügel, Lori A Burns, T Daniel Crawford, William Dawson, Wibe Albert de Jong, Claudia Draxl, Claudia Filippi, Luigi Genovese, Paolo Giannozzi, Niranjan Govind, Sharon Hammes-Schiffer, Jeff R Hammond, Benjamin Hourahine, Anubhav Jain, Yosuke Kanai, Paul R C Kent, Ask Hjorth Larsen, Susi Lehtola, Xiaosong Li, Roland Lindh, Satoshi Maeda, Nancy Makri, Jonathan Moussa, Takahito Nakajima, Jessica A Nash, Micael J T Oliveira, Pansy D Patel, Giovanni Pizzi, Geoffrey Pourtois, Benjamin P Pritchard, Eran Rabani, Markus Reiher, Lucia Reining, Xinguo Ren, Mariana Rossi, H Bernhard Schlegel, Nicola Seriani, Lyudmila V Slipchenko, Alexander Thom, Edward F Valeev, Benoit Van Troeye, Lucas Visscher, Vojtěch Vlček, Hans-Joachim Werner, David B Williams-Young, Theresa Windus
16418
Reposted by Jan Hermann
Derk Kooi @derkkooi.bsky.social · 16/11/2024
In this post-truth era we simply decide it to be correct. Repeat after me: CCSD(T) is the gold standard of quantum chemistry, CCSD(T) is the gold standard of quantum chemistry, CCSD(T) is the gold standard of quantum chemistry.
132
Jan Hermann @hrmnn.net · 16/11/2024
Hot take—CCSD(T) is about as ab-initio, or as empirical, as PBE. Why the particular formulation of perturbative triples in CCSD(T) should work so well in practice is about as clear as why the particular choice of constraints in PBE does. In both cases empirical evidence is the primary argument.
13321