Notice: The reproducibility variables underlying each score are classified using an automated LLM-based pipeline, validated against a manually labeled dataset. LLM-based classification introduces uncertainty and potential bias; scores should be interpreted as estimates. Full accuracy metrics and methodology are described in Coakley et alK. L. Coakley, T. Snelleman, H. Hoos, and O. E. Gundersen, "The embrace of open science: An analysis of a decade of AI research and 56 800 conference papers," Under Review, 2026..
FlowDAS: A Stochastic Interpolant-based Framework for Data Assimilation
Authors: Siyi Chen, Yixuan Jia, Qing Qu, He Sun, Jeffrey A. Fessler
NeurIPS 2025 | Venue PDF | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Experiments on Lorenz-63, Navier Stokes super-resolution/sparse-observation scenarios, and large-scale weather forecasting where dynamics are partly or wholly unknown show that Flow DAS surpasses model-driven methods, neural operators, and score-based baselines in accuracy and physical plausibility. |
| Researcher Affiliation | Academia | Siyi Chen University of Michigan EMAIL Yixuan Jia University of Michigan EMAIL Qing Qu University of Michigan EMAIL He Sun Peking University EMAIL Jeffrey A. Fessler University of Michigan EMAIL |
| Pseudocode | Yes | Algorithm 1 Training Algorithm 2 Inference |
| Open Source Code | Yes | Our implementation is available at https://github.com/umjiayx/Flow DAS. |
| Open Datasets | Yes | Experiments on Lorenz-63, Navier Stokes super-resolution/sparse-observation scenarios, and large-scale weather forecasting... using the Storm EVent Imagery and Radar (SEVIR) dataset [54]. |
| Dataset Splits | Yes | Dataset and experiments We generate 1,024 independent trajectories, each containing 1,024 states, and split the data into training (80%), validation (10%), and evaluation (10%) sets. |
| Hardware Specification | Yes | Table S.10: Comparison of model size, total training time, and inference time on the weather forecasting task using a single A100 GPU. |
| Software Dependencies | No | The paper does not provide specific software versions for libraries or environments used, beyond naming models (e.g., FNO, Transolver) and optimizers (Adam, Adam W). |
| Experiment Setup | Yes | For this low-dimensional problem, we use a fully connected neural network to approximate the drift term in stochastic interpolants... The model is optimized using collected trajectories x0:K, minimizing an empirical loss between predicted and target velocities... The model is trained using the Adam optimizer with a base learning rate of 0.005, along with a linear rate scheduler. Training is conducted for 5000 epochs. ... We use N = 500 in our experiments. |