REVIEW 12 cited by
Benchmarking Autoregressive Conditional Diffusion Models for Turbulent Flow Simulation
Not yet reviewed by Pith; the record is open.
This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.
SPECIMEN: schema-true, not a live event
T0 review · schema-true
One-sentence machine reading of the paper's core claim.
pith:XXXXXXXX · record.json · timestamp
Signed reviews
read the original abstract
Simulating turbulent flows is crucial for a wide range of applications, and machine learning-based solvers are gaining increasing relevance. However, achieving temporal stability when generalizing to longer rollout horizons remains a persistent challenge for learned PDE solvers. In this work, we analyze if fully data-driven fluid solvers that utilize an autoregressive rollout based on conditional diffusion models are a viable option to address this challenge. We investigate accuracy, posterior sampling, spectral behavior, and temporal stability, while requiring that methods generalize to flow parameters beyond the training regime. To quantitatively and qualitatively benchmark the performance of various flow prediction approaches, three challenging 2D scenarios including incompressible and transonic flows, as well as isotropic turbulence are employed. We find that even simple diffusion-based approaches can outperform multiple established flow prediction methods in terms of accuracy and temporal stability, while being on par with state-of-the-art stabilization techniques like unrolling at training time. Such traditional architectures are superior in terms of inference speed, however, the probabilistic nature of diffusion approaches allows for inferring multiple predictions that align with the statistics of the underlying physics. Overall, our benchmark contains three carefully chosen data sets that are suitable for probabilistic evaluation alongside various established flow prediction architectures.
Forward citations
Cited by 12 Pith papers
-
Round-Trip Consistency: Bidirectional Diffusion Models Can Predict Their Own Rollout Errors
A model's forward-then-backward round-trip error predicts, without ensembles or ground truth, how large its actual rollout error will be.
-
Point-wise Diffusion Models for Physical Systems with Shape Variations: Application to Spatio-temporal and Large-scale system
A point-wise diffusion transformer predicts spatio-temporal physical fields on arbitrary meshes and point clouds, claiming up to 200x faster inference and better accuracy than image-based diffusion surrogates.
-
PDE-Transformer: Efficient and Versatile Transformers for Physics Simulations
PDE-Transformer, a diffusion-transformer variant with shifted-window attention, multi-scale token processing, and per-channel tokens, outperforms leading transformer and operator baselines for PDE surrogate modeling a...
-
Autoregressive regularized score-based diffusion models for multi-scenarios fluid flow prediction
A regularized autoregressive score-based diffusion model predicts turbulent flows across multiple scenarios, with the variance-preserving SDE formulation performing best.
-
STFlow: Data-Coupled Flow Matching for Geometric Trajectory Simulation
A flow-matching model that builds its starting noise from random walks matched to the observed motion produces more accurate trajectory predictions with only 5 integration steps.
-
FLEX: A Backbone for Diffusion-Based Modeling of Spatio-temporal Physical Systems
A hybrid U-Net and Transformer diffusion backbone trained in residual space with a velocity parametrization outperforms baselines on 2D turbulence super-resolution and forecasting, and generalizes zero-shot to unseen ...
-
Conditional flow matching for generative modeling of near-wall turbulence with quantified uncertainty
A method using conditional flow matching with a SWAG-trained forward model reconstructs near-wall turbulent velocity fields from wall measurements, producing an ensemble with quantified uncertainty.
-
Benchmarking Multi-fidelity Neural Operators on Complex PDE Problems with Non-trivial Fidelity Differences
Across four PDE benchmarks, transfer learning from low-fidelity weights is the only multi-fidelity neural operator strategy that consistently outperforms the high-fidelity-only baseline, while direct LF-input methods ...
-
Integrating Fourier Neural Operator with Diffusion Model for Autoregressive Predictions of Three-dimensional Turbulence
DiAFNO, an implicit adaptive Fourier neural operator used as the denoiser inside an EDM diffusion model, gives more accurate autoregressive predictions of 3D turbulence than EDM or dynamic Smagorinsky LES.
-
Inferring processes within dynamic forest models using hybrid modeling
The abstract claims a hybrid gap-model plus neural-network approach, FINN, improves forest growth inference and forecasting, but the manuscript body is an unrelated diffusion-model paper, so the abstract's claims are ...
-
Predicting Change, Not States: An Alternate Framework for Neural PDE Surrogates
Predicting the temporal derivative and integrating it with an ODE solver improves accuracy and stability of neural PDE surrogates compared with direct next-state prediction.
-
Evaluation of Neural Surrogates for Physical Modelling Synthesis of Nonlinear Elastic Plates
On a Berger plate benchmark, state-of-the-art neural surrogates fail in long autoregressive rollouts, and time-domain error metrics miss the resulting spectral errors.
Discussion (0). Continue with ORCID to comment.