Stars
Physics-informed neural networks for highly compressible flows 🧠🌊
[CVPR2025] Breaking the Low-Rank Dilemma of Linear Attention
Next Generation Experimental Tracking for Machine Learning Operations
Möller–Trumbore ray-triangle intersection algorithm.
[NeurIPS 2025] Accelerated Neural Simulations of Three-Dimensional Turbulence at Scale
Build high-quality Laplace matrices on meshes and point clouds in Python. Implements [Sharp & Crane SGP 2020].
ResiDual: Transformer with Dual Residual Connections, https://arxiv.org/abs/2304.14802
Discrete wavelet transform (DWT) via lifting in PyTorch
pyGeo provides geometric design variables and constraints suitable for gradient-based optimization.
High-performance 2D solver for Kelvin-Helmholtz instability in incompressible flows, featuring Numba JIT compilation for 10-50x speedup and multi-core parallelization. Includes predefined scenarios…
A FEniCS Project-based library for simulating thin structures
Implementation of 'Physics-Informed Neural Networks for Shell Structures' (European Journal of Mechanics A)
An inverse design framework combining SDF representations with a classifier-free guided diffusion model for one-shot synthesis of functional microstructures. Conditioned on target stress-strain cur…
Unofficial JAX implementation of the SOAP optimizer (https://arxiv.org/abs/2409.11321)
pretraining dataset for 2D PDE foundation models
HAET: Hierarchical Attention Erwin Transolver is a hybrid neural architecture that combines physically-aware spatial decomposition with hierarchical attention for efficient and accurate learning on…
Mixture-of-Recursions: Learning Dynamic Recursive Depths for Adaptive Token-Level Computation (NeurIPS 2025)
Einstein Fields official implementation in JAX.
Official PyTorch Implementation of "Scalable Diffusion Models with Transformers"
We propose S2GPT-PINN, a sparse and small model for solving parametric partial differential equations (PDEs).