Scientific Machine Learning
PulseAugur coverage of Scientific Machine Learning — every cluster mentioning Scientific Machine Learning across labs, papers, and developer communities, ranked by signal.
5 day(s) with sentiment data
-
Production-Ready Scientific ML: Beyond Test Scores
The concept of "production-ready" in Scientific Machine Learning (SML) extends beyond mere test scores. It encompasses the ability to reproduce, validate, deploy, monitor, and ultimately trust a model. This distinction …
-
New Transformer Architecture Enhances Operator Learning on Complex Geometries
Researchers have introduced ArGEnT, a novel geometry-encoded Transformer designed for operator learning on arbitrary geometries. This framework decouples geometry encoding from query-point evaluation, enabling mesh-inde…
-
New TIDE benchmark dataset aims to advance 3D turbulence ML research
Researchers have introduced TIDE, a new benchmark dataset designed to advance scientific machine learning in the field of 3D turbulence. TIDE provides a large-scale dataset of 3D incompressible turbulence simulations, f…
-
New research outlines SGD preconditioner design for stability and noise reduction
A new research paper published on arXiv details design criteria for stochastic gradient descent (SGD) preconditioners, focusing on local conditioning, noise floors, and basin stability. The paper derives bounds where co…
-
New framework bridges hybrid models with neuro-symbolic AI
Researchers have developed a new framework called Hybrid-to-NeSy (H2N) that bridges hybrid mechanistic/data-driven models with neuro-symbolic AI. This approach translates hybrid modeling designs into a neuro-symbolic in…
-
Label-free training method for neural surrogates in fluid dynamics
Researchers have developed a novel method for training neural surrogates for thermo-fluid field predictions, utilizing a label-free approach based on minimizing finite-volume method (FVM) residuals. This technique, appl…
-
New PA-SciML workflow verifies physics compliance in agentic SciML discovery
Researchers have introduced Physics-Audited Agentic SciML (PA-SciML), a new workflow designed to enhance the reliability of scientific machine learning (SciML) models discovered by large language model (LLM) agents. Thi…
-
New Zeroth-Order Deep Learning Method Tackles High-Dimensional PDEs
Researchers have developed a novel zeroth-order deep learning method to tackle high-dimensional partial differential equations (PDEs) with unknown coefficients, a common challenge in scientific machine learning and cont…
-
Scientific Machine Learning advances fluid dynamics simulation
A recent chapter reviews advancements in Scientific Machine Learning (SciML) for simulating complex fluid flow and transport phenomena. It highlights methods like Dynamic Mode Decomposition and Physics-Informed Neural N…
-
Scientific Machine Learning advances fluid dynamics modeling · 2 sources tracked
This chapter explores advancements in Scientific Machine Learning (SciML) for simulating complex fluid flow and transport phenomena. It details methods like Singular Value Decomposition, Dynamic Mode Decomposition, Phys…
-
Paper links neural operators to differential equations for better generalization
A new paper explores the relationship between traditional differential equation models and modern data-driven approaches like neural operators. It argues that many modeling strategies share a common structure, differing…
-
New 'instrumented data' concept advances scientific machine learning
Researchers have introduced a new concept called "instrumented data" for scientific machine learning, aiming to overcome limitations in current data types. This approach embeds the mechanistic model, its uncertainty, an…
-
New method boosts PDE pre-training with adaptive operator transformation
Researchers have developed AOT-POT, a novel method for pre-training neural operators on diverse partial differential equation (PDE) datasets. This approach transforms complex solution operators into simpler, aligned for…