Symmetry in the Wild: The Role of Equivariance in Neural Fluid Surrogates
The paper discusses the role of equivariance in neural fluid surrogates, which can significantly speed up computational fluid dynamics simulations. It highlights the conditions under which equivariance can enhance generalization, particularly in automotive aerodynamics and blood flow applications. The study introduces a new model, the Anchored-Branched Geometric Algebra Transformer, which demonstrates the benefits of equivariance in specific scenarios while noting its limitations in others.
- ▪Neural surrogates can accelerate computational fluid dynamics simulations by orders of magnitude.
- ▪The study investigates how equivariance affects generalization in neural CFD surrogates across various tasks.
- ▪The Anchored-Branched Geometric Algebra Transformer model integrates scalability and symmetry preservation.
Opening excerpt (first ~120 words) tap to expand
Computer Science > Machine Learning arXiv:2605.18816 (cs) [Submitted on 12 May 2026] Title:Symmetry in the Wild: The Role of Equivariance in Neural Fluid Surrogates Authors:Patryk Rygiel, Julian Suk, Kak Khee Yeung, Christoph Brune, Jelmer M. Wolterink View a PDF of the paper titled Symmetry in the Wild: The Role of Equivariance in Neural Fluid Surrogates, by Patryk Rygiel and 4 other authors View PDF HTML (experimental) Abstract:Neural surrogates enable orders-of-magnitude acceleration of computational fluid dynamics (CFD) simulations, with the potential to transform engineering and healthcare workflows.
…
Excerpt limited to ~120 words for fair-use compliance. The full article is at arXiv cs.AI.
Discussion
0 commentsMore from arXiv cs.AI
