arXiv subjects

Florian Schäfer

Publications and source records attributed to Florian Schäfer.

At least 19 recordsLinked to original sources

arXiv2026

Information geometric regularization for computing sensitivities of flows with shocks

Computing (adjoint) sensitivities of flows with shocks is a longstanding problem in computational fluid dynamics. The spurious sensitivities due to shock sensors and limiters frequently force practitioners to accept the errors incurred by "freezing" limiters and shock sensors. The recently proposed information geometric regularization (IGR) is an inviscid, PDE-based regularization of the compressible Euler equations that replaces shocks with smooth profiles without damping fine-scale structures. This work derives the forward and adjoint sensitivities for the IGR system with periodic boundary conditions and demonstrates their convergence, under grid refinement, to the sensitivities obtained by finite differences or automatic differentiation through the forward solve.

math.NA↗

arXiv2026

Shocks without shock capturing: Information geometric regularization of finite volume methods for Navier--Stokes-like problems

Shock waves in high-speed fluid dynamics produce near-discontinuities in the fluid momentum, density, and energy. Most contemporary works use artificial viscosity or limiters as numerical mitigation of the Gibbs--Runge oscillations that result from traditional numerics. These approaches face a delicate balance in achieving sufficiently regular solutions without dissipating fine-scale features, such as turbulence or acoustics. Recent work by Cao and Schäfer introduces information geometric regularization (IGR), the first inviscid regularization method for fluid dynamics. IGR replaces shock singularities with smooth profiles of adjustable width, without dissipating fine-scale features. This work provides a strategy for the practical use of IGR in finite-volume-based numerical methods. We illustrate its performance on canonical test problems and compare it against established approaches based on limiters and Riemann solvers. Results show that the finite volume IGR approach recovers the expected solutions in all cases. Across canonical benchmarks, IGR achieves accuracy competitive with WENO and LAD shock-capturing schemes in both smooth and discontinuous flow regimes. The IGR approach is computationally light, with meaningfully fewer memory accesses and arithmetic operations per time step.

cs.CE↗

arXiv2026

Discontinuous Galerkin Semidiscretization of the Information Geometric Regularized Compressible Euler Equations

Shock stabilization in compressible Euler flows remains a central challenge for high-order numerical methods. Existing shock-capturing approaches, including limiters, artificial viscosity, and reconstruction-based methods, involve tradeoffs between robustness, accuracy, preservation of fine-scale flow features, and computational complexity. In this work, we develop a discontinuous Galerkin (DG) discretization of the information geometric regularization (IGR) framework introduced by Cao and Schäfer for the compressible Euler equations. The method regularizes shocks at the PDE level by augmenting the Euler equations with the entropic pressure $Σ$, obtained from an auxiliary elliptic equation. Within the DG formulation, the regularization enters only through the augmented pressure $P+Σ$ in the Euler fluxes, preserving the conservative structure of the discretization while using a common approximation space for both the hyperbolic and elliptic equations. Numerical experiments spanning one and two-dimensional benchmark problems show the proposed formulation stabilizes shocks without shock-capturing limiters or artificial viscosity, although positivity-preserving methods may still be required when the density or pressure approaches zero. Compared with a characteristic TVB-limited DG formulation, the IGR-DG method resolves increasingly finer-scale flow features as the polynomial order is increased while maintaining stable shock resolution. The entropic pressure remains localized to regions of strong compression with minimal activation in smooth regions of the flow, providing selective PDE-level regularization while preserving the underlying solution elsewhere.

math.NA↗

arXiv2026

Shock solutions for the one-dimensional information geometric regularization of compressible flow

The information geometric regularization (IGR) is an inviscid regularization of the compressible Euler equations that alters the geometry of Lagrangian characteristics to prevent trajectories from crossing in finite time. Previous work on IGR established global strong solutions in one dimension, explored thermodynamic effects of the model, and enabled large-scale simulations of compressible flow. However, a fundamental question that remains is how this regularization alters the structure and regularity of a shock-like solution. We prove existence, uniqueness modulo translation, and regularity of transonic compressive IGR shock profiles in one spatial dimension. The analysis applies to the full thermodynamic compressible Euler--IGR model with a general equation of state, subject to mild convexity hypotheses. A traveling-wave ansatz reduces the Euler--IGR equations to a degenerate second-order scalar equation for the density profile. At the sonic crossing, the elliptic coefficient degenerates: the density profile remains continuous, but its derivative diverges. The profile is a classical solution away from this single point, while at the degeneracy it retains quantified Hölder and Sobolev regularity. We also analyze the vanishing-regularization limit, showing that the shock width scales like $\sqrt{α}$ and that the IGR profiles converge to the entropy-admissible Euler shock.

math.AP↗

arXiv2026

Information geometric regularization of unidimensional pressureless Euler equations yields global strong solutions

Partial differential equations describing compressible fluids are prone to the formation of shock singularities, arising from faster upstream fluid particles catching up to slower, downstream ones. In geometric terms, this causes the deformation map to leave the manifold of diffeomorphisms. Information geometric regularization addresses this issue by changing the manifold geometry to make it geodesically complete. Empirical evidence suggests that this results in smooth solutions without adding artificial viscosity. This work makes a first step towards understanding this phenomenon rigorously, in the setting of the unidimensional pressureless Euler equations. It shows that their information geometric regularization has smooth global solutions. By establishing $Γ$-convergence of its variational description, it proves convergence of these solutions to entropy solutions of the nominal problem, in the limit of vanishing regularization parameter. A consequence of these results is that manifolds of unidimensional diffeomorphisms with information geometric regularization are geodesically complete.

math.AP↗

arXiv2026

Influence of Chemistry and Topography on the Wettability of Copper

To understand the complex interplay of topography and surface chemistry in wetting, fundamental studies investigating both parameters are needed. Due to the sensitivity of wetting to miniscule changes in one of the parameters it is imperative to precisely control the experimental approach. A profound understanding of their influence on wetting facilitates a tailored design of surfaces with unique functionality. We present a multi-step study: The influence of surface chemistry is analyzed by determining the adsorption of volatile carbonous species (A) and by sputter deposition of metallic copper and copper oxides on flat copper substrates (B). A precise surface topography is created by laser processing. Isotropic topography is created by ps laser processing (C), and hierarchical anisotropic line patterns are produced by direct laser interference patterning (DLIP) with different pulse durations (D). Our results reveal that the long-term wetting response of polished copper surfaces stabilizes with time despite ongoing accumulation of hydrocarbons and is dominated by this adsorption layer over the oxide state of the substrate (Cu, CuO, Cu2O). The surfaces' wetting response can be precisely tuned by tailoring the topography via laser processing. The sub-pattern morphology of primary line-like patterns showed great impact on the static contact angle, wetting anisotropy, and water adhesion. An increased roughness inside the pattern valleys combined with a minor roughness on the peaks favors air-inclusions, isotropic hydrophobicity, and low water adhesion. Increasing the aspect ratio showed to enhance air-inclusions and hydrophobicity despite increased peak roughness while time dependent wetting transitions were observed.

cond-mat.mtrl-sci↗

arXiv2026

Information geometric regularization of the barotropic Euler equation

Shock waves in gas dynamics feature jump discontinuities that hinder numerical simulations. Viscous regularizations are prone to excessive dissipation of fine-scale structures. In this work, we propose the first inviscid regularization of the multidimensional Euler equation based on ideas from semidefinite programming, information geometry, geometric hydrodynamics, and nonlinear elasticity. The Lagrangian flow maps of Euler solutions are a dynamical system on the manifold of diffeomorphisms. We observe that shock formation arises from the manifold's geodesic incompleteness. Our regularization embeds it into an ambient space equipped with the information geometry of the logarithmic barrier function. Thus, the diffeomorphism manifold inherits a geodesically complete geometry. The resulting regularized conservation law replaces shocks with smooth profiles without affecting oscillatory structures. One and two-dimensional numerical experiments show its practical potential to enable higher-order methods without explicit shock capturing. While we focus on the barotropic Euler equations for concreteness and simplicity of exposition, our regularization easily extends to more general Euler and Navier-Stokes-type equations. Our approach regularizes the Wasserstein geometry of the mass density with its information geometry. The former captures the natural trajectories of physical particles and the latter that of statistical estimators. Information geometric regularization accounts for the mass density's dual nature as a statistical/computational tool summarizing the motion of physical particles. Thus, our work is a starting point for information geometric mechanics that views solutions of continuum mechanical PDEs as parameters of statistical models for unresolved scales and uses their information geometry to evolve them in time.

math.NA↗

arXiv2026

Möbius inversion and the iterated bootstrap

Estimating nonlinear functionals of probability distributions from samples is a fundamental statistical problem. The "plug-in" estimator obtained by applying the target functional to the empirical distribution of samples is biased. Resampling methods such as the bootstrap derive artificial datasets from the original one by resampling. Comparing the outcome of the plug-in estimator in the original and resampled datasets allows estimating and thus correcting the bias. In the asymptotic setting, iterations of this procedure attain an arbitrarily high order of bias correction, but finite sample results are scarce. This work develops a new theoretical understanding of bootstrap bias correction by viewing it as an iterative linear solver for the combinatorial operation of Möbius inversion. It sharply characterizes the regime of linear convergence of the bootstrap bias reduction for moment polynomials. It uses these results to show its superalgebraic convergence rate for band-limited functionals. Finally, it derives a modified bootstrap iteration enabling the unbiased estimation of unknown order-$m$ moment polynomials in $m$ bootstrap iterations.

math.ST↗

arXiv2025

Hamiltonian Information Geometric Regularization of the Compressible Euler Equations

The recently proposed information geometric regularization (IGR) was the first inviscid regularization of the multi-dimensional compressible Euler equations, which enabled the simulation of realistic compressible fluid models at an unprecedented scale. However, the thermodynamic effects of this regularization have not yet been understood in a principled manner. To achieve a proper understanding of the thermodynamic aspects of the IGR, we decompose the regularization into its conservative dynamics, framed as a Hamiltonian subsystem, and its dissipative dynamics. In so doing, we further introduce two more models to compare to IGR, the Hamiltonian regularized Euler (HRE) model, which is the first multi-dimensional, non-dispersive Hamiltonian regularization of the compressible Euler equations with energy, as well as the Hamiltonian IGR (HIGR) model, which modifies the dissipation used in the IGR model to instead utilize a metriplectic dissipative force. Despite having many attractive features, the HRE and HIGR models exhibit notable defects in numerical tests on colliding shock problems, which preclude their use as computational tools without further study of dissipative weak solutions to these models. Additionally, our analysis presents new results on the IGR model itself, including its ability to conserve acoustic waves, as well as local energy transport laws and entropy production rates. By separating the conservative and dissipative dynamics of the IGR, our hope is that subsequent of analysis of the IGR model can benefit from this natural decomposition, such as, for example, rigorous proofs of strong solutions for multi-dimensional IGR for the compressible Euler system with thermodynamics.

math-ph↗

arXiv2025

A Bootstrap Perspective on Stochastic Gradient Descent

Machine learning models trained with \emph{stochastic} gradient descent (SGD) can generalize better than those trained with deterministic gradient descent (GD). In this work, we study SGD's impact on generalization through the lens of the statistical bootstrap: SGD uses gradient variability under batch sampling as a proxy for solution variability under the randomness of the data collection process. We use empirical results and theoretical analysis to substantiate this claim. In idealized experiments on empirical risk minimization, we show that SGD is drawn to parameter choices that are robust under resampling and thus avoids spurious solutions even if they lie in wider and deeper minima of the training loss. We prove rigorously that by implicitly regularizing the trace of the gradient covariance matrix, SGD controls the algorithmic variability. This regularization leads to solutions that are less sensitive to sampling noise, thereby improving generalization. Numerical experiments on neural network training show that explicitly incorporating the estimate of the algorithmic variability as a regularizer improves test performance. This fact supports our claim that bootstrap estimation underpins SGD's generalization advantages.

cs.LG↗

arXiv2025

Simulating many-engine spacecraft: Exceeding 1 quadrillion degrees of freedom via information geometric regularization

We present an optimized implementation of the recently proposed information geometric regularization (IGR) for unprecedented scale simulation of compressible fluid flows applied to multi-engine spacecraft boosters. We improve upon state-of-the-art computational fluid dynamics (CFD) techniques along computational cost, memory footprint, and energy-to-solution metrics. Unified memory on coupled CPU--GPU or APU platforms increases problem size with negligible overhead. Mixed half/single-precision storage and computation on well-conditioned numerics is used. We simulate flow at 200 trillion grid points and 1 quadrillion degrees of freedom, exceeding the current record by a factor of 20. A factor of 4 wall-time speedup is achieved over optimized baselines. Ideal weak scaling is seen on OLCF Frontier, LLNL El Capitan, and CSCS Alps using the full systems. Strong scaling is near ideal at extreme conditions, including 80% efficiency on CSCS Alps with an 8-node baseline and stretching to the full system.

physics.comp-ph↗

arXiv2025

Maximum likelihood discretization of the transport equation

The transport of positive quantities underlies countless physical processes, including fluid, gas, and plasma dynamics. Discretizing the associated partial differential equations with Galerkin methods can result in spurious nonpositivity of solutions. We observe that these methods amount to performing statistical inference using the method of moments (MoM) and that the loss of positivity arises from MoM's susceptibility to producing estimates inconsistent with the observed data. We overcome this problem by replacing MoM with maximum likelihood estimation, introducing $\textit{maximum likelihood discretization} $(MLD). In the continuous limit, MLD simplifies to the Fisher-Rao Galerkin (FRG) semidiscretization, which replaces the $L^2$ inner product in Galerkin projection with the Fisher-Rao metric of probability distributions. We show empirically that FRG preserves positivity. We prove rigorously that it yields error bounds in the Kullback-Leibler divergence.

math.NA↗

arXiv2025

Sparse inverse Cholesky factorization of dense kernel matrices by greedy conditional selection

Dense kernel matrices resulting from pairwise evaluations of a kernel function arise naturally in machine learning and statistics. Previous work in constructing sparse approximate inverse Cholesky factors of such matrices by minimizing Kullback-Leibler divergence recovers the Vecchia approximation for Gaussian processes. These methods rely only on the geometry of the evaluation points to construct the sparsity pattern. In this work, we instead construct the sparsity pattern by leveraging a greedy selection algorithm that maximizes mutual information with target points, conditional on all points previously selected. For selecting $k$ points out of $N$, the naive time complexity is $\mathcal{O}(N k^4)$, but by maintaining a partial Cholesky factor we reduce this to $\mathcal{O}(N k^2)$. Furthermore, for multiple ($m$) targets we achieve a time complexity of $\mathcal{O}(N k^2 + N m^2 + m^3)$, which is maintained in the setting of aggregated Cholesky factorization where a selected point need not condition every target. We apply the selection algorithm to image classification and recovery of sparse Cholesky factors. By minimizing Kullback-Leibler divergence, we apply the algorithm to Cholesky factorization, Gaussian process regression, and preconditioning with the conjugate gradient, improving over $k$-nearest neighbors selection.

stat.CO↗

arXiv2025

Optimal Krylov On Average

We propose an adaptive randomized truncation estimator for Krylov subspace methods that optimizes the trade-off between the solution variance and the computational cost, while remaining unbiased. The estimator solves a constrained optimization problem to compute the truncation probabilities on the fly, with minimal computational overhead. The problem has a closed-form solution when the improvement of the deterministic algorithm satisfies a diminishing returns property. We prove that obtaining the optimal adaptive truncation distribution is impossible in the general case. Without the diminishing return condition, our estimator provides a suboptimal but still unbiased solution. We present experimental results in GP hyperparameter training and competitive physics-informed neural networks problem to demonstrate the effectiveness of our approach.

math.NA↗

arXiv2024

Sparse Cholesky Factorization for Solving Nonlinear PDEs via Gaussian Processes

In recent years, there has been widespread adoption of machine learning-based approaches to automate the solving of partial differential equations (PDEs). Among these approaches, Gaussian processes (GPs) and kernel methods have garnered considerable interest due to their flexibility, robust theoretical guarantees, and close ties to traditional methods. They can transform the solving of general nonlinear PDEs into solving quadratic optimization problems with nonlinear, PDE-induced constraints. However, the complexity bottleneck lies in computing with dense kernel matrices obtained from pointwise evaluations of the covariance kernel, and its \textit{partial derivatives}, a result of the PDE constraint and for which fast algorithms are scarce. The primary goal of this paper is to provide a near-linear complexity algorithm for working with such kernel matrices. We present a sparse Cholesky factorization algorithm for these matrices based on the near-sparsity of the Cholesky factor under a novel ordering of pointwise and derivative measurements. The near-sparsity is rigorously justified by directly connecting the factor to GP regression and exponential decay of basis functions in numerical homogenization. We then employ the Vecchia approximation of GPs, which is optimal in the Kullback-Leibler divergence, to compute the approximate factor. This enables us to compute $ε$-approximate inverse Cholesky factors of the kernel matrices with complexity $O(N\log^d(N/ε))$ in space and $O(N\log^{2d}(N/ε))$ in time. We integrate sparse Cholesky factorizations into optimization algorithms to obtain fast solvers of the nonlinear PDE. We numerically illustrate our algorithm's near-linear space/time complexity for a broad class of nonlinear PDEs such as the nonlinear elliptic, Burgers, and Monge-Ampère equations.

math.NA↗

arXiv2024

Asymptotic properties of Vecchia approximation for Gaussian processes

Vecchia approximation has been widely used to accurately scale Gaussian-process (GP) inference to large datasets, by expressing the joint density as a product of conditional densities with small conditioning sets. We study fixed-domain asymptotic properties of Vecchia-based GP inference for a large class of covariance functions (including Matérn covariances) with boundary conditioning. In this setting, we establish that consistency and asymptotic normality of maximum exact-likelihood estimators imply those of maximum Vecchia-likelihood estimators, and that exact GP prediction can be approximated accurately by Vecchia GP prediction, given that the size of conditioning sets grows polylogarithmically with the data size. Hence, Vecchia-based inference with quasilinear complexity is asymptotically equivalent to exact GP inference with cubic complexity. This also provides a general new result on the screening effect. Our findings are illustrated by numerical experiments, which also show that Vecchia approximation can be more accurate than alternative approaches such as covariance tapering and reduced-rank approximations.

math.ST↗

arXiv2024

InRank: Incremental Low-Rank Learning

The theory of greedy low-rank learning (GLRL) aims to explain the impressive generalization capabilities of deep learning. It proves that stochastic gradient-based training implicitly regularizes neural networks towards low-rank solutions through a gradual increase of the rank during training. However, there is a gap between theory and practice since GLRL requires an infinitesimal initialization of the weights, which is not practical due to the fact that it is a saddle point. In this work, we remove the assumption of infinitesimal initialization by focusing on cumulative weight updates. We prove the cumulative weight updates follow an incremental low-rank trajectory for arbitrary orthogonal initialization of weights in a three-layer linear network. Empirically, we demonstrate that our theory holds on a broad range of neural networks (e.g., transformers) and standard training algorithms (e.g., SGD, Adam). However, existing training algorithms do not exploit the low-rank property to improve computational efficiency as the networks are not parameterized in low-rank. To remedy this, we design a new training algorithm Incremental Low-Rank Learning (InRank), which explicitly expresses cumulative weight updates as low-rank matrices while incrementally augmenting their ranks during training. We evaluate InRank on GPT-2, and our results indicate that InRank achieves comparable prediction performance as the full-rank counterpart while requiring at most 33% of the total ranks throughout training. We also propose an efficient version of InRank that achieves a reduction of 37% in total training time and 36% in model size when training GPT-medium on WikiText-103 from scratch.

cs.LG↗

arXiv2023

Targeted computation of nonlocal closure operators via an adjoint-based macroscopic forcing method

Reynolds-averaged Navier--Stokes (RANS) closure must be sensitive to the flow physics, including nonlocality and anisotropy of the effective eddy viscosity. Recent approaches used forced direct numerical simulations to probe these effects, including the macroscopic forcing method (MFM) of Mani and Park ($\textit{Phys. Rev. Fluids}$ $\textbf{6}$, 054607 (2021)) and the Green's function approach of Hamba ($\textit{Phys. Fluids}$ $\textbf{17}$, 115102 (2005)). The resulting nonlocal and anisotropic eddy viscosities are exact and relate Reynolds stresses to mean velocity gradients at all locations. They can be used to inform RANS models of the sensitivity to the mean velocity gradient and the suitability of local and isotropic approximations. However, these brute-force approaches are expensive. They force the mean velocity gradient at each point in the averaged space and measure the Reynolds stress response, requiring a separate simulation for each mean velocity gradient location. Thus, computing the eddy viscosity requires as many simulations as degrees of freedom in the averaged space, which can be cost-prohibitive for problems with many degrees of freedom. In this work, we develop an adjoint-based MFM to obtain the eddy viscosity at a given Reynolds stress location using a single simulation. This approach recovers the Reynolds stress dependence at a location of interest, such as a separation point or near a wall, on the mean velocity gradient at all locations. We demonstrate using adjoint MFM to compute the eddy viscosity for a specified wall-normal location in an incompressible turbulent channel flow using one simulation. In contrast, a brute-force approach for the same problem requires $N=144$ simulations (the number of grid points in the non-averaged coordinate direction). We show that a local approximation for the eddy viscosity would have been inappropriate.

physics.flu-dyn↗