Spectrally regularized compression in latent flow matching raises retained deep-dissipation spectral power from 20% to 79% in generated turbulence on a 256^2 DNS dataset at Re_f ≈ 2250.
hub
Smith, Ayya Alieva, Qing Wang, Michael P
15 Pith papers cite this work, alongside 46 external citations. Polarity classification is still indexing.
15
Pith papers citing it
46
external citations · OpenAlex
hub tools
citation-role summary
background 1
method 1
citation-polarity summary
representative citing papers
PDEInvBench supplies a new benchmark dataset for PDE inverse problems and shows that two-stage supervised-plus-test-time training, PDE derivative inputs, and diverse initial conditions improve neural network accuracy on parameter recovery.
JAX-FVM is a differentiable, entropy-stable finite volume solver for 2D compressible Euler/Navier-Stokes equations on unstructured meshes, built in JAX.
Mosaic is a benchmark suite evaluating 14 differentiable PDE solvers across fluids, structures, and heat transfer, showing large variations in cost and conditioning but similar convergence behavior.
PI-MFA optimizes tensor-product B-spline control points to balance data fidelity against PDE residuals, producing physically consistent continuous flow fields.
A new JAX library provides GPU-accelerated AMG-preconditioned sparse linear solves with reverse-mode automatic differentiation, JIT, and MPI support.
A label-conditioned drifting model in VAE latent space matches diffusion accuracy for flow surrogates while running two orders of magnitude faster, with a spatial variant for unseen geometries.
Mamba-Assisted Closure (MAC) trains a Mamba sequence model on resolved trajectories to predict non-Markovian closures and couples it with reduced-order equations, outperforming Markovian, GRU, and Wilks baselines on Burgers' and Lorenz '96 systems.
ARC-STAR reduces velocity rollout error by at least 36x over raw Poseidon across all tested regime cells via auditable global and local correction stages on five flow benchmarks.
ShardTensor is a domain-parallelism system for SciML that enables flexible scaling of extreme-resolution spatial datasets by removing the constraint of batch size one per device.
LASER couples a latent world model with a GRPO-trained RL policy to adaptively reposition sensors for sparse continuum field reconstruction, outperforming fixed and offline-optimized placement strategies across PDE and real-world benchmarks.
QIML uses a quantum-trained Q-Prior to enhance classical autoregressive predictions of spatiotemporal chaos, improving accuracy by up to 17.25% and full-spectrum fidelity by up to 29.36% while enabling stable forecasts for 3D turbulent channel flow.
ELGIN is a graph-based physics-informed surrogate model that predicts carrier flow and polydisperse particle motion in dental aerosol scenarios, achieving lower tracking errors and 37x speedup versus full OpenFOAM CFD in a preliminary single-case test.
MENO restores multi-scale structure in neural-operator PDE surrogates via one-step improved MeanFlow, claiming up to 2× better power-spectrum accuracy and up to 14× faster inference than DDIM enhancement.
Porting AI-accelerated CFD model training to IPU-POD16 yields 34% data-feeding speedup and scales throughput to 2805 samples/s on 16 IPUs despite inter-IPU communication limits.
citing papers explorer
-
Spectrally Regularized Latent Flow Matching for Turbulence Generation
Spectrally regularized compression in latent flow matching raises retained deep-dissipation spectral power from 20% to 79% in generated turbulence on a 256^2 DNS dataset at Re_f ≈ 2250.
-
PDEInvBench: A Comprehensive Dataset and Design Space Exploration of Neural Networks for PDE Inverse Problems
PDEInvBench supplies a new benchmark dataset for PDE inverse problems and shows that two-stage supervised-plus-test-time training, PDE derivative inputs, and diverse initial conditions improve neural network accuracy on parameter recovery.
-
JAX-FVM: A differentiable, entropy-stable finite volume solver on unstructured meshes for compressible flows
JAX-FVM is a differentiable, entropy-stable finite volume solver for 2D compressible Euler/Navier-Stokes equations on unstructured meshes, built in JAX.
-
Mosaic: A Benchmark Suite for Differentiable Physics Solvers
Mosaic is a benchmark suite evaluating 14 differentiable PDE solvers across fluids, structures, and heat transfer, showing large variations in cost and conditioning but similar convergence behavior.
-
A Physics-Informed B-Spline Framework for Continuous Approximation of Flow Data
PI-MFA optimizes tensor-product B-spline control points to balance data fidelity against PDE residuals, producing physically consistent continuous flow fields.
-
JAX-AMG: A GPU-Accelerated Differentiable Sparse Linear Solver Library for JAX
A new JAX library provides GPU-accelerated AMG-preconditioned sparse linear solves with reverse-mode automatic differentiation, JIT, and MPI support.
-
Drifting Models for Surrogate Flow Modeling
A label-conditioned drifting model in VAE latent space matches diffusion accuracy for flow surrogates while running two orders of magnitude faster, with a spatial variant for unseen geometries.
-
Mamba-Assisted Non-Markovian Closure for Reduced-Order Modeling
Mamba-Assisted Closure (MAC) trains a Mamba sequence model on resolved trajectories to predict non-Markovian closures and couples it with reduced-order equations, outperforming Markovian, GRU, and Wilks baselines on Burgers' and Lorenz '96 systems.
-
ARC-STAR: Auditable Post-Hoc Correction for PDE Foundation Models
ARC-STAR reduces velocity rollout error by at least 36x over raw Poseidon across all tested regime cells via auditable global and local correction stages on five flow benchmarks.
-
ShardTensor: Domain Parallelism for Scientific Machine Learning
ShardTensor is a domain-parallelism system for SciML that enables flexible scaling of extreme-resolution spatial datasets by removing the constraint of batch size one per device.
-
LASER: Learning Active Sensing for Continuum Field Reconstruction
LASER couples a latent world model with a GRPO-trained RL policy to adaptively reposition sensors for sparse continuum field reconstruction, outperforming fixed and offline-optimized placement strategies across PDE and real-world benchmarks.
-
Quantum-Informed Machine Learning for Predicting Spatiotemporal Chaos with Practical Quantum Advantage
QIML uses a quantum-trained Q-Prior to enhance classical autoregressive predictions of spatiotemporal chaos, improving accuracy by up to 17.25% and full-spectrum fidelity by up to 29.36% while enabling stable forecasts for 3D turbulent channel flow.
-
Physics-Informed Graph Neural Network Surrogates for Turbulent Nanoparticle Dispersion in Dental Clinical Environments
ELGIN is a graph-based physics-informed surrogate model that predicts carrier flow and polydisperse particle motion in dental aerosol scenarios, achieving lower tracking errors and 37x speedup versus full OpenFOAM CFD in a preliminary single-case test.
-
MENO: MeanFlow-Enhanced Neural Operators for Dynamical Systems
MENO restores multi-scale structure in neural-operator PDE surrogates via one-step improved MeanFlow, claiming up to 2× better power-spectrum accuracy and up to 14× faster inference than DDIM enhancement.
-
Adaptation of AI-accelerated CFD Simulations to the IPU platform
Porting AI-accelerated CFD model training to IPU-POD16 yields 34% data-feeding speedup and scales throughput to 2805 samples/s on 16 IPUs despite inter-IPU communication limits.