mlsanityData to 14 Oct 2026

DISCO: learning to DISCover an evolution Operator for multi-physics-agnostic prediction

arXiv 2504.19496 ↗First posted 28 Apr 2025cs.LG cs.AI stat.MLcited by 21

Not on the map. Why?

Abstract (arXiv)

We address the problem of predicting the next state of a dynamical system governed by unknown temporal partial differential equations (PDEs) using only a short trajectory. While standard transformers provide a natural black-box solution to this task, the presence of a well-structured evolution operator in the data suggests a more tailored and efficient approach. Specifically, when the PDE is fully known, classical numerical solvers can evolve the state accurately with only a few parameters. Building on this observation, we introduce DISCO, a model that uses a large hypernetwork to process a short trajectory and generate the parameters of a much smaller operator network, which then predicts the next state through time integration. Our framework decouples dynamics estimation (i.e., DISCovering an evolution operator from a short trajectory) from state prediction (i.e., evolving this operator). Experiments show that pretraining our model on diverse physics datasets achieves state-of-the-art performance while requiring significantly fewer epochs. Moreover, it generalizes well and remains competitive when fine-tuned on downstream tasks.

Key facts. Cited by 21 papers in the map (data through 14 Oct 2026). Builds on 46 of its 68 references that are in the map. First posted on arXiv 28 Apr 2025.

Builds on

46 of 68 references.

References · 46

  1. 2017Attention Is All You Needcited by 60,361
  2. 2020Language Models are Few-Shot Learnerscited by 32,419
  3. 2019PyTorch: An Imperative Style, High-Performance Deep Learning Librarycited by 20,305
  4. 2015U-Net: Convolutional Networks for Biomedical Image Segmentationcited by 19,713
  5. 2024The Llama 3 Herd of Modelscited by 14,715
  6. 2017Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networkscited by 4,897
  7. 2018Neural Ordinary Differential Equationscited by 3,779
  8. 2016Gaussian Error Linear Units (GELUs)cited by 3,502
  9. 2020Fourier Neural Operator for Parametric Partial Differential Equationscited by 2,908
  10. 2018UMAP: Uniform Manifold Approximation and Projection for Dimension Reductioncited by 2,898
  11. 2018Group Normalizationcited by 1,524
  12. 2016HyperNetworkscited by 1,027
34 more
  1. 2020Learning Mesh-Based Simulation with Graph Networkscited by 868
  2. 2021Machine learning accelerated computational fluid dynamicscited by 471
  3. 2022PDEBENCH: An Extensive Benchmark for Scientific Machine Learningcited by 413
  4. 2022Message Passing Neural PDE Solverscited by 386
  5. 2017PDE-Net: Learning PDEs from Datacited by 311
  6. 2022On Neural Differential Equationscited by 268
  7. 2024Poseidon: Efficient Foundation Models for PDEscited by 200
  8. 2018Deep Neural Networks Motivated by Partial Differential Equationscited by 187
  9. 2022Adan: Adaptive Nesterov Momentum Algorithm for Faster Optimizing Deep Modelscited by 174
  10. 2024DPOT: Auto-Regressive Denoising Operator Transformer for Large-Scale PDE Pre-Trainingcited by 151
  11. 2024The Well: a Large-Scale Collection of Diverse Physics Simulations for Machine Learningcited by 145
  12. 2023In-Context Operator Learning with Data Prompts for Differential Equation Problemscited by 101
  13. 2023Multiple Physics Pretraining for Physical Surrogate Modelscited by 98
  14. 2024ClimODE: Climate and Weather Forecasting with Physics-informed Neural ODEscited by 82
  15. 2020MeshfreeFlowNet: A Physics-Constrained Deep Continuous Space-Time Super-Resolution Frameworkcited by 78
  16. 2019Learning Neural PDE Solvers with Convergence Guaranteescited by 77
  17. 2023A Mathematical Guide to Operator Learningcited by 76
  18. 2022Rediscovering orbital mechanics with machine learningcited by 71
  19. 2023Benchmarking Autoregressive Conditional Diffusion Models for Turbulent Flow Simulationcited by 71
  20. 2024Neural Operators with Localized Integral and Differential Kernelscited by 62
  21. 2023Towards Stability of Autoregressive Neural Operatorscited by 59
  22. 2022Generalizing to New Physical Systems via Context-Informed Dynamics Modelcited by 50
  23. 2022Efficient Learning of Mesh-Based Physical Simulation with BSMS-GNNcited by 45
  24. 2022Multi-resolution partial differential equations preserved learning framework for spatiotemporal dynamicscited by 44
  25. 2021Meta-Learning Dynamics Forecasting Using Task Inferencecited by 38
  26. 2024PDE Generalization of In-Context Operator Networks: A Study on 1D Scalar Nonlinear Conservation Lawscited by 31
  27. 2023Self-Supervised Learning with Lie Symmetries for Partial Differential Equationscited by 29
  28. 2024VICON: Vision In-Context Operator Networks for Multi-Physics Fluid Dynamics Predictioncited by 25
  29. 2020Generative ODE Modeling with Known Unknownscited by 24
  30. 2024Zebra: In-Context Generative Pretraining for Solving Parametric PDEscited by 16
  31. 2023Deciphering and integrating invariants for neural operator learning with various physical mechanismscited by 15
  32. 2023Interpretable Meta-Learning of Physical Systemscited by 11
  33. 2024GEPS: Boosting Generalization in Parametric PDE Neural Solvers through Adaptive Conditioningcited by 7
  34. 2021Frame-independent vector-cloud neural network for nonlocal constitutive modeling on arbitrary gridscited by 6

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 7; 2026: 14.

21 papers.

Sort:

By area:

  1. 2025Walrus: A Cross-Domain Foundation Model for Continuum Dynamicscited by 36
  2. 2025Lost in Latent Space: An Empirical Study of Latent Diffusion Models for Physics Emulationcited by 19
  3. 2025MORPH: PDE Foundation Models with Arbitrary Data Modalitycited by 14
  4. 2026Representation Learning for Spatiotemporal Physical Systemscited by 11
  5. 2026Test-time Generalization for Physics through Neural Operator Splittingcited by 10
  6. 2025HyPINO: Multi-Physics Neural Operators via HyperPINNs and the Method of Manufactured Solutionscited by 8
  7. 2025Towards a Physics Foundation Modelcited by 8
  8. 2026Probabilistic Retrofitting of Learned Simulatorscited by 7
  9. 2025Weight-Space Linear Recurrent Neural Networkscited by 5
  10. 2026Emergent Transfer of a Physics Foundation Model from Simulation to Laboratory Turbulencecited by 4
  11. 2025Learning where to learn: Training data distribution optimization for scientific machine learningcited by 3
  12. 2026HyCOP: Hybrid Composition Operators for Interpretable Learning of PDEscited by 2
9 more
  1. 2026Iterative Refinement Neural Operators are Learned Fixed-Point Solvers: A Principled Approach to Spectral Bias Mitigationcited by 2
  2. 2026Factorized Neural Implicit DMD for Parametric Dynamicscited by 1
  3. 2026MVNN: A Measure-Valued Neural Network for Learning McKean-Vlasov Dynamics from Particle Datacited by 1
  4. 2026Posterior-First Neural PDE Simulation: Inferring Hidden Problem State from a Single Fieldcited by 0
  5. 2026A Robust In-Context Model for Conservation Laws: Injecting Context into Flux Neural Operators via Recurrent Vision Transformerscited by 0
  6. 2026Eradicating Negative Transfer in Multi-Physics Foundation Models via Sparse Mixture-of-Experts Routingcited by 0
  7. 2026Breakeven complexity: A new perspective on neural partial differential equation solverscited by 0
  8. 2026PHASE: Multi-Regime Modeling of Incompressible Magnetohydrodynamicscited by 0
  9. 2026Cross-Domain Pretraining for Steady-State Neural CFD Surrogatescited by 0