FlowMM: Generating Materials with Riemannian Flow Matching

Benjamin Kurt Miller, Ricky T. Q. Chen, Anuroop Sriram, Brandon M Wood
Proceedings of the 41st International Conference on Machine Learning, PMLR 235:35664-35686, 2024.

Abstract

Crystalline materials are a fundamental component in next-generation technologies, yet modeling their distribution presents unique computational challenges. Of the plausible arrangements of atoms in a periodic lattice only a vanishingly small percentage are thermodynamically stable, which is a key indicator of the materials that can be experimentally realized. Two fundamental tasks in this area are to (a) predict the stable crystal structure of a known composition of elements and (b) propose novel compositions along with their stable structures. We present FlowMM, a pair of generative models that achieve state-of-the-art performance on both tasks while being more efficient and more flexible than competing methods. We extend Riemannian Flow Matching to suit the symmetries inherent to crystals: translation, rotation, permutation, and periodic boundary conditions. Our framework enables the freedom to choose the flow base distributions, drastically simplifying the problem of learning crystal structures compared with diffusion models. In addition to standard benchmarks, we validate FlowMM’s generated structures with quantum chemistry calculations, demonstrating that it is $\sim$3x more efficient, in terms of integration steps, at finding stable materials compared to previous open methods.

Cite this Paper

BibTeX
@InProceedings{pmlr-v235-miller24a, title = {{F}low{MM}: Generating Materials with {R}iemannian Flow Matching}, author = {Miller, Benjamin Kurt and Chen, Ricky T. Q. and Sriram, Anuroop and Wood, Brandon M}, booktitle = {Proceedings of the 41st International Conference on Machine Learning}, pages = {35664--35686}, year = {2024}, editor = {Salakhutdinov, Ruslan and Kolter, Zico and Heller, Katherine and Weller, Adrian and Oliver, Nuria and Scarlett, Jonathan and Berkenkamp, Felix}, volume = {235}, series = {Proceedings of Machine Learning Research}, month = {21--27 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v235/main/assets/miller24a/miller24a.pdf}, url = {https://proceedings.mlr.press/v235/miller24a.html}, abstract = {Crystalline materials are a fundamental component in next-generation technologies, yet modeling their distribution presents unique computational challenges. Of the plausible arrangements of atoms in a periodic lattice only a vanishingly small percentage are thermodynamically stable, which is a key indicator of the materials that can be experimentally realized. Two fundamental tasks in this area are to (a) predict the stable crystal structure of a known composition of elements and (b) propose novel compositions along with their stable structures. We present FlowMM, a pair of generative models that achieve state-of-the-art performance on both tasks while being more efficient and more flexible than competing methods. We extend Riemannian Flow Matching to suit the symmetries inherent to crystals: translation, rotation, permutation, and periodic boundary conditions. Our framework enables the freedom to choose the flow base distributions, drastically simplifying the problem of learning crystal structures compared with diffusion models. In addition to standard benchmarks, we validate FlowMM’s generated structures with quantum chemistry calculations, demonstrating that it is $\sim$3x more efficient, in terms of integration steps, at finding stable materials compared to previous open methods.} }
Endnote
%0 Conference Paper %T FlowMM: Generating Materials with Riemannian Flow Matching %A Benjamin Kurt Miller %A Ricky T. Q. Chen %A Anuroop Sriram %A Brandon M Wood %B Proceedings of the 41st International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2024 %E Ruslan Salakhutdinov %E Zico Kolter %E Katherine Heller %E Adrian Weller %E Nuria Oliver %E Jonathan Scarlett %E Felix Berkenkamp %F pmlr-v235-miller24a %I PMLR %P 35664--35686 %U https://proceedings.mlr.press/v235/miller24a.html %V 235 %X Crystalline materials are a fundamental component in next-generation technologies, yet modeling their distribution presents unique computational challenges. Of the plausible arrangements of atoms in a periodic lattice only a vanishingly small percentage are thermodynamically stable, which is a key indicator of the materials that can be experimentally realized. Two fundamental tasks in this area are to (a) predict the stable crystal structure of a known composition of elements and (b) propose novel compositions along with their stable structures. We present FlowMM, a pair of generative models that achieve state-of-the-art performance on both tasks while being more efficient and more flexible than competing methods. We extend Riemannian Flow Matching to suit the symmetries inherent to crystals: translation, rotation, permutation, and periodic boundary conditions. Our framework enables the freedom to choose the flow base distributions, drastically simplifying the problem of learning crystal structures compared with diffusion models. In addition to standard benchmarks, we validate FlowMM’s generated structures with quantum chemistry calculations, demonstrating that it is $\sim$3x more efficient, in terms of integration steps, at finding stable materials compared to previous open methods.
APA
Miller, B.K., Chen, R.T.Q., Sriram, A. & Wood, B.M.. (2024). FlowMM: Generating Materials with Riemannian Flow Matching. Proceedings of the 41st International Conference on Machine Learning, in Proceedings of Machine Learning Research 235:35664-35686 Available from https://proceedings.mlr.press/v235/miller24a.html.

Related Material