✨ new work: "Truncated automatic sparse differentiation for MLIPs" with @adrianhill.de & @micheleceriotti.bsky.social has arrived on the arXiv. In a 🥥: we exploit that MLIP Hessians are sparse and decay with distance with a "truncated" variant of automatic sparse differentiation. See figure. 👇

Show Post

We have released ShiftML4! ShiftML4 comes with built-in molecular corrections, elevating the chemical shielding predictions to hybrid-DFT level quality. Try it yourself on shiftml.org, or from the PyPi package.

Show Post

Time to learn the action 🏃 ! Today I'm sharing a paper, just published on @physrevlett.bsky.social, showing how to construct a symplectic #machinelearning predictor of classical mechanics by learning the Hamilton-Jacobi action.

Show Post

I don’t have the budget for hundreds of GPUs or even a backend, so I had to make mine use your local resources 🤓 peterspackman.github.io/mlip.cpp/

Show Post

How well can you predict mechanical properties (elastic tensors) for molecular crystals *just* looking at pairwise interaction energies? Surprisingly well! With a big asterisk... Read all the ways it fails and where it works in JCTC now: pubs.acs.org/doi/full/10....

Show Post