iiinova — where your digital self keeps evolving

Adam Foster

@adamefoster.bsky.social

87 followers 50 following 8 posts

Sr Researcher at Microsoft Research AI for Science Former Oxford Stats PhD in Bayesian Experimental Design

Posts Replies Media

Reposted by Adam Foster

Frank Noe @franknoe.bsky.social · 10/07/2025

BioEmu now published in @science.org !! What is BioEmu? Check out this video: youtu.be/LStKhWcL0VE?...

science.org

Scalable emulation of protein equilibrium ensembles with generative deep learning

Following the sequence and structure revolutions, predicting functionally relevant protein structure changes at scale remains an outstanding challenge. We introduce BioEmu, a deep learning system that...

69128

Adam Foster @adamefoster.bsky.social · 26/06/2025

Digging into the model we found intriguing behaviour, such as the unsupervised discovery by the model of ‘core’ electron orbitals for second row atoms. This has been a fascinating project to be a part of! Check out the preprint for more details and results. (8/n)

120

Adam Foster @adamefoster.bsky.social · 26/06/2025

We also saw very strong results confirming the experimental activation energy of a Diels-Alder reaction, and significantly outperforming earlier transferable QMC approaches (7/n)

120

Adam Foster @adamefoster.bsky.social · 26/06/2025

We scaled this idea up and pushed it to work on strongly correlated systems. On a cost/error plot, we find that Orbformer is on or ahead of the Pareto frontier formed by traditional multireference methods, a first for deep QMC. (6/n)

110

Adam Foster @adamefoster.bsky.social · 26/06/2025

To get cost down we make use of amortization: solving a single minimization problem with a more complex network that represents multiple wavefunctions simultaneously (5/n)

110

Adam Foster @adamefoster.bsky.social · 26/06/2025

Describing strongly correlated quantum systems remains a major challenge in quantum chemistry. Deep QMC offer a potential solution, but at a huge computational cost. (4/n)

110

Adam Foster @adamefoster.bsky.social · 26/06/2025

For a short ML focused introduction to this paper, I wrote this blog post ae-foster.github.io/posts/2025/0... (3/n)

ae-foster.github.io

A machine learning introduction to Orbformer

We have just released Orbformer, a foundation model of molecular wavefunctions! This is a project I have been working on over the last two-and-a-bit years as part of an amazing team. This blog post te...

130

Adam Foster @adamefoster.bsky.social · 26/06/2025

Huge thanks to Zeno Schätzle, P Bernát Szabó, @lixuecheng-ai4chem.bsky.social @jonkhler.argmin.xyz @n-gao.bsky.social Jiawei Li, Gino Cassella @franknoe.bsky.social @jan.hermann.name (2/n)

130

Adam Foster @adamefoster.bsky.social · 26/06/2025

I am very happy to share Orbformer, a foundation model for wavefunctions using deep QMC that offers a route to tackle strongly correlated quantum states! arxiv.org/abs/2506.19960 (1/n)

arxiv.org

An ab initio foundation model of wavefunctions that accurately describes chemical bond breaking

Reliable description of bond breaking remains a major challenge for quantum chemistry due to the multireferential character of the electronic structure in dissociating species. Multireferential method...

1297

Reposted by Adam Foster

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