iiinova — where your digital self keeps evolving

iiinova is part of the Atmosphere — an open, public network where you own your identity.

Aditi Krishnapriyan

@ask1729.bsky.social

2.1K followers 154 following 9 posts

Assistant Professor at UC Berkeley

Posts Replies Media

Aditi Krishnapriyan @ask1729.bsky.social · 13/03/2025

6/ The distilled MLFFs are much faster to run than the original large-scale MLFF: not everyone has the GPU resources to use big models and many scientists only care about studying specific systems (w/ the correct physics!). This is a way to get the best of all worlds!

110 1

Aditi Krishnapriyan @ask1729.bsky.social · 13/03/2025

5/ We can also balance training at scale efficiently (often w/ minimal constraints) with distilling the correct physics into the small MLFF at test time: e.g., taking energy gradients to get conservative forces, and ensuring energy conservation for molecular dynamics.

100 1

Aditi Krishnapriyan @ask1729.bsky.social · 13/03/2025

4/ Smaller, specialized MLFFs distilled from the large-scale model are more accurate than training from scratch on the same subset of data: the representations from the large-scale model help boost performance, while the smaller models are much faster to run

100 1

Aditi Krishnapriyan @ask1729.bsky.social · 13/03/2025

3/ We formulate our distillation procedure as the smaller MLFF is trained to match Hessians of the energy predictions of the large-scale model (using subsampling methods to improve efficiency). This works better than distillation methods to try to match features.

110 1

Aditi Krishnapriyan @ask1729.bsky.social · 13/03/2025

1/ Machine learning force fields are hot right now 🔥: models are getting bigger + being trained on more data. But how do we balance size, speed, and specificity? We introduce a method for doing model distillation on large-scale MLFFs into fast, specialized MLFFs! More details below: #ICLR2025

1112 11