A Stability-Aware Frozen Euler Autoencoder for Physics-Informed Tracking in Continuum Mechanics (SAFE-PIT-CM)

• URL: http://arxiv.org/abs/2603.13280v2
• Date: Wed, 18 Mar 2026 19:34:03 GMT
• Title: A Stability-Aware Frozen Euler Autoencoder for Physics-Informed Tracking in Continuum Mechanics (SAFE-PIT-CM)
• Abstract summary: We introduce SAFE-PIT-CM, a Stability-Aware Frozen Euler autoencoder for Physics-Informed Tracking in Continuum Mechanics.<n>The architecture embeds a frozen, differentiable PDE solver inside an autoencoder.<n>The model also supports test-time training (TTT): learning alpha from a single simulation with no pre-training, using only the physics loss as supervision.
• Score: 0.0
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: We introduce SAFE-PIT-CM, a Stability-Aware Frozen Euler autoencoder for Physics-Informed Tracking in Continuum Mechanics. The architecture embeds a frozen, differentiable PDE solver inside an autoencoder: a convolutional encoder maps each video frame to a latent field; the SAFE operator propagates it forward via sub-stepped finite differences; and a decoder reconstructs the video. Backpropagation through the frozen operator yields gradients that directly supervise an attention-based estimator for the diffusion coefficient alpha (alpha = D > 0), requiring no ground-truth labels. The SAFE operator is the central contribution. In practice, temporal snapshots are saved at intervals far coarser than the simulation time step. A single forward Euler step at this coarse interval violates the von Neumann stability condition, causing alpha to collapse to an unphysical value. Sub-stepping the frozen stencil restores the original temporal resolution and numerical stability, enabling accurate parameter recovery across the full physically relevant range. SAFE-PIT-CM is demonstrated on the diffusion equation, covering thermal diffusion in engineering metals. The model also supports test-time training (TTT): learning alpha from a single simulation with no pre-training, using only the physics loss as supervision. TTT achieves accuracy comparable to pre-trained inference. The architecture generalises to any PDE admitting a convolutional finite-difference discretisation and is inherently explainable -- every prediction traces to a physical coefficient and step-by-step PDE propagation.

Related papers

• Temperature Field Reconstruction of Tungsten Monoblock Divertor on EAST using Physics-aware Neural Operator Transformer [20.873931353548574]
  We propose a Physics-aware Neural Operator Transformer (PNOT) to characterize the evolution of the divertor temperature field.<n>Inspired by physics-aware attention, we further develop a physics-aware neural operator module to aggregate query points.<n> Experimental results show that these physical constraints improve prediction accuracy while preserving physical consistency.
  arXiv  Detail & Related papers  (2026-06-30T12:33:06Z)
• Deformable Gaussian Occupancy: Decoupling Rigid and Nonrigid Motion with Factorized Distillation [81.44955493339835]
  DeGO is a deformable Gaussian occupancy framework that unifies decoupled Gaussian deformation with factorized 4D foundation-model distillation.<n> Experiments on the Occ3D-NuScenes benchmark demonstrate that our method achieves state-of-the-art performance under weak supervision.
  arXiv  Detail & Related papers  (2026-05-27T15:09:55Z)
• Physics from Video: Identifiability of Time-Invariant Second-Order ODEs under Minimal Trajectory Conditions [60.2784641643737]
  Bridging the gap between visual realism and physical understanding is a core challenge for video-based world models.<n>We study the structural identifiability of continuous-time physical laws from raw pixels.<n>We prove that a level-set slope-coverage condition ensures the learned latent space is locally affine to the true physical state.
  arXiv  Detail & Related papers  (2026-05-27T13:22:29Z)
• NPSolver: Neural Poisson Solver with Iterative Physics Supervision [13.109423311684134]
  textscNPr is a neural Poisson solver trained without solution labels.<n>textscNPr utilizes a small number of preconditioned conjugate gradient steps to refine its own predictions.<n>textscBA-Transolver architecture explicitly separates interior and boundary tokenization.
  arXiv  Detail & Related papers  (2026-05-25T12:33:52Z)
• Sparse Autoencoders as a Steering Basis for Phase Synchronization in Graph-Based CFD Surrogates [6.396157607535812]
  Graph-based surrogate models provide fast alternatives to high-fidelity CFD solvers, but their opaque latent spaces and limited controllability restrict use in safety-critical settings.<n>We ask whether phase drift can instead be corrected post hoc by manipulating the latent space of a frozen surrogate.<n>We propose a phase-steering framework for pretrained graph-based CFD models that combines the right representation with the right intervention mechanism.
  arXiv  Detail & Related papers  (2026-03-28T07:03:44Z)
• Stabilizing Physics-Informed Consistency Models via Structure-Preserving Training [7.031010831953522]
  We propose a physics-informed consistency modeling framework for solving partial differential equations (PDEs)<n>We identify a key stability challenge in physics-constrained consistency training, where PDE residuals can drive the model toward trivial or degenerate solutions.<n>We introduce a structure-preserving two-stage training strategy that decouples distribution learning from physics enforcement.
  arXiv  Detail & Related papers  (2026-02-10T00:40:19Z)
• On Exact Editing of Flow-Based Diffusion Models [97.0633397035926]
  We propose Conditioned Velocity Correction (CVC) to reformulate flow-based editing as a distribution transformation problem driven by a known source prior.<n>CVC rethinks the role of velocity in inter-distribution transformation by introducing a dual-perspective velocity conversion mechanism.<n>We show that CVC consistently achieves superior fidelity, better semantic alignment, and more reliable editing behavior across diverse tasks.
  arXiv  Detail & Related papers  (2025-12-30T06:29:20Z)
• CFO: Learning Continuous-Time PDE Dynamics via Flow-Matched Neural Operators [9.273461312644345]
  Continuous Flow Operator (CFO) learns continuous-time PDE dynamics without the computational burden of standard continuous approaches, e.g., neural ODE.<n>CFO fits temporal splines to trajectory data, using finite-difference estimates of time derivatives at knots to construct probability paths whose velocities closely approximate the true PDE dynamics.<n>A neural operator is then trained via flow matching to predict these analytic velocity fields.
  arXiv  Detail & Related papers  (2025-12-04T22:33:29Z)
• INC: An Indirect Neural Corrector for Auto-Regressive Hybrid PDE Solvers [61.84396402100827]
  We propose the Indirect Neural Corrector ($mathrmINC$), which integrates learned corrections into the governing equations.<n>$mathrmINC$ reduces the error amplification on the order of $t-1 + L$, where $t$ is the timestep and $L$ the Lipschitz constant.<n>We test $mathrmINC$ in extensive benchmarks, covering numerous differentiable solvers, neural backbones, and test cases ranging from a 1D chaotic system to 3D turbulence.
  arXiv  Detail & Related papers  (2025-11-16T20:14:28Z)
• EntroPE: Entropy-Guided Dynamic Patch Encoder for Time Series Forecasting [50.794700596484894]
  We propose EntroPE (Entropy-Guided Dynamic Patch), a novel, temporally informed framework that dynamically detects transition points via conditional entropy.<n>This preserves temporal structure while retaining the computational benefits of patching.<n> Experiments across long-term forecasting benchmarks demonstrate that EntroPE improves both accuracy and efficiency.
  arXiv  Detail & Related papers  (2025-09-30T12:09:56Z)
• PICT -- A Differentiable, GPU-Accelerated Multi-Block PISO Solver for Simulation-Coupled Learning Tasks in Fluid Dynamics [62.93137406343609]
  We present our fluid simulator PICT, a differentiable pressure-implicit solver coded in PyTorch with Graphics-processing-unit (GPU) support.<n>We first verify the accuracy of both the forward simulation and our derived gradients in various established benchmarks.<n>We show that the gradients provided by our solver can be used to learn complicated turbulence models in 2D and 3D.
  arXiv  Detail & Related papers  (2025-05-22T17:55:10Z)
• A finite element-based physics-informed operator learning framework for spatiotemporal partial differential equations on arbitrary domains [33.7054351451505]
  We propose a novel finite element-based physics operator learning framework that allows for predicting dynamics governed by partial differential equations (PDEs) The proposed operator learning framework takes a temperature field at the current time step as input and predicts a temperature field at the next time step. Networks successfully predict the temperature evolution over time for any initial temperature field at high accuracy compared to the FEM solution.
  arXiv  Detail & Related papers  (2024-05-21T02:41:40Z)
• FC-PINO: High Precision Physics-Informed Neural Operators via Fourier Continuation [60.706803227003995]
  We introduce the FC-PINO (Fourier-Continuation-based PINO) architecture which extends the accuracy and efficiency of PINO to non-periodic and non-smooth PDEs.<n>We demonstrate that standard PINO struggles to solve non-periodic and non-smooth PDEs with high precision, across challenging benchmarks.<n>In contrast, the proposed FC-PINO provides accurate, robust, and scalable solutions, substantially outperforming PINO alternatives.
  arXiv  Detail & Related papers  (2022-11-29T06:37:54Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.