Joe Abbott

And also Guillaume talking (very soon at 10:45) in Room 1 about all things metatensor and metatomic - making it easier to develop and use ML models for atomistic simulations 🔥 #psik2025

27.08.2025 08:34 — 👍 0    🔁 0    💬 0    📌 0

If you are at the #psik2025 and want to know more about the #metatensor ecosystem, don't miss @luthaf.bsky.social talk tomorrow morning 9:45 in room 1

26.08.2025 21:31 — 👍 1    🔁 1    💬 0    📌 1

Come and see my poster "On the importance of symmetry constraints for learning equivariant quantum mechanical properties" at Psi-k this lunchtime, poster B5.01!

27.08.2025 08:30 — 👍 1    🔁 0    💬 0    📌 0

A cartoon explaining how mild finite-temperature conditions induce disorder and dynamical reconstruction on the surfaces of lithium thiophosphates

📢 Now out on @physrevx.bsky.social energy, journals.aps.org/prxenergy/ab... from 🧑‍🚀 @dtisi.bsky.social and Hanna Türk, our #PET -powered study of the dynamic reconstruction of LPS surfaces, and how it affects their structure, stability and reactivity.

27.08.2025 06:54 — 👍 8    🔁 4    💬 1    📌 0

metatensor logo

metatomic logo

🚨 #machinelearning for #compchem goodies from our 🧑‍🚀 team incoming! After years of work it's time to share. Go check arxiv.org/abs/2508.15704 and/or metatensor.org to learn about #metatensor and #metatomic. What they are, what they do, why you should use them for all of your atomistic ML projects 🔍.

22.08.2025 07:40 — 👍 10    🔁 7    💬 1    📌 2

Scheme of the GNN architecture of the FlashMD method.

📢 Running molecular dynamics with time steps up to 64fs for any atomistic system, from Al(110) to Ala2? Thanks to 🧑‍🚀 Filippo Bigi and Sanggyu Chong, with some help from Agustinus Kristiadis, this is not as crazy as it sounds. Let us briefly introduce FlashMD⚡ arxiv.org/html/2505.19...

27.05.2025 07:02 — 👍 37    🔁 12    💬 1    📌 1

Polar plot showing the errors of several machine-learning potential of different test sets. Smaller is better here!

Plots showing the evaluation time per atom for several machine-learning potentials as a function of the number of atoms in a simulation. Smaller is better

📢 PET-MAD has just landed! 📢 What if I told you that you can match & improve the accuracy of other "universal" #machinelearning potentials training on fewer than 100k atomic structures? And be *faster* with an unconstrained architecture that is conservative with tiny symmetry breaking? Sounds like 🧑‍🚀

19.03.2025 07:23 — 👍 28    🔁 9    💬 1    📌 3