arXiv subjects

Phil-Alexander Hofmann

Publications and source records attributed to Phil-Alexander Hofmann.

9 recordsLinked to original sources

arXiv2026

Analysis of the Ill-Conditioning of the Discrete Inverse Laplace transform in Monte Carlo Simulations of Quantum Many-Body Systems

Analytic continuation of imaginary-time quantum Monte Carlo data to real-frequency spectra requires the inversion of a severely ill-posed two-sided Laplace transform and arises naturally in quantum many-body calculations of dynamic properties. In this work, we distinguish the intrinsic ill-posedness of the continuous inverse two-sided Laplace problem from the conditioning of its finite-dimensional discretization. For equidistant sampling and reconstruction grids, we express the discrete problem via a diagonally scaled monomial Vandermonde matrix with exponentially distributed nodes. Imposing the physical detailed-balance symmetry transforms the discretization into a diagonally scaled Chebyshev-Vandermonde system. Exploiting these structures, we derive explicit lower and upper bounds on the condition numbers in terms of the physical and discretization parameters. For the unconstrained discretization, our bounds reveal super-exponential growth of the condition number with the reconstruction dimension, which cannot be removed by increasing the number of imaginary-time samples alone. Detailed-balance substantially improves the conditioning, especially in the practically relevant pre-asymptotic regime, although the asymptotic super-exponential dependence remains. In both cases, our bounds identify a low-dimensional regime in which the super-exponential contribution is suppressed and stable reconstruction remains feasible. These results provide a mathematical explanation for the effectiveness of low-dimensional spectral representations and detailed-balance in analytic continuation. While they do not remove the fundamental ill-posedness, they substantially mitigate the ill-conditioning of its finite-dimensional discretization, delay the onset of its catastrophic super-exponential growth and thereby enlarge the range of numerically accessible reconstructions.

math.NA↗

arXiv2026

Mitigating Numerical Stiffness in Least-Squares Formulations of Elliptic PDEs for Physics-Informed Neural Networks

We present theoretical insights into $H^{-1}$ residual loss formulations of physics-informed neural networks (PINNs) for learning solutions of partial differential equations (PDEs). Standard PINN formulations use a multi-term loss functional consisting of interior and boundary loss terms that are based on $L^2$-residuals and discretized as mean square errors (MSE). Imbalanced magnitudes of these terms cause numerical stiffness phenomena, resulting in ill-conditioning and slow convergence. In this work, we analyze discretizations of the $H^{-1}$-norm that are used in the context of elliptic PDEs with arbitrary, nonzero Dirichlet boundary conditions. We prove that these $H^{-1}$ discretizations rebalance the PDE loss, improve conditioning, and mitigate stiffness effects compared with the standard MSE discretization. We validate our theoretic results through operator-level experiments with randomly sampled residuals and PINN experiments for the Poisson and stationary incompressible Navier-Stokes equations. These experiments confirm the numerical effectiveness of the proposed rebalancing for elliptic PDEs and, more broadly, for problems with elliptic behavior.

math.NA↗

arXiv2026

Discovering a well-conditioned analytic continuation problem via dictionary learning

Many fields of physics use quantum Monte Carlo (QMC) simulations to simulate quantum systems in imaginary-time $\tau$ and estimate imaginary-time correlation functions (ITCF). However, extracting dynamic $\omega$-dependent quantities from ITCFs is a notoriously difficult task, known as analytic continuation (AC), that amounts to solving an exponentially ill-conditioned inverse problem. Within the AC literature, there are competing stochastic and regularized approaches, as well as an emerging collection of works using parameterized models like neural networks. Here we transcend the traditional divides between the communities, introducing the regularized stochastic optimization method (RSOM). This method reformulates AC as a dictionary learning problem, discovering a sparse dictionary to represent the solution. Our approach is motivated by the astounding results dictionary learning has produced in many scientific fields. Remarkably, RSOM discovers a sparse dictionary that maps an ill-conditioned inverse problem to a low-dimensional problem that is well-conditioned. We demonstrate that the method yields competitive results for common synthetic test problems as well as for authentic QMC data from the finite temperature electron gas. This work exposes that a dictionary exists within all stochastic and regularized methods and that dictionary learning provides a new angle of attack for future AC methods.

physics.comp-ph↗

arXiv2025

Interpolation in Polynomial Spaces of p-Degree

We recently introduced the Fast Newton Transform (FNT), an hierarchical algorithm for performing multivariate Newton interpolation in arbitrary downward closed polynomial spaces of spatial dimension $m$. Here, we analyze the FNT in the context of a specific family of downward closed sets $A_{m,n,p}$, defined as all multi-indices with $\ell^p$ norm less than $n$ with $p \in [0,\infty]$. The FNT performs with time complexity $\mathcal{O}(|A_{m,n,p}|mn)$ on the induced downward closed polynomial spaces $\Pi_{m,n,p}$. We show that the $\Pi_{m,n,p}$ choice compared to the tensor product spaces $\Pi_{m,n,\infty}$, reduces time complexity by a factor of $\rho_{m,n,p}$, decaying super exponentially with spatial dimension when $m \lesssim n^p$. We showcase the efficiency of the FNT by computing activity scores in sensitivity analysis.

math.NA↗

arXiv2025

Accelerating Multivariate Newton Interpolation in Downward Closed Polynomial Spaces

We introduce the fast Newton transform (FNT), a multivariate Newton interpolation algorithm for downward closed polynomial spaces in quasi-tensorial grids. The FNT computes the Newton coefficients directly, without relying on embeddings into enclosing tensor-product spaces. For a downward closed index set $A \subset \mathbb N_0^m$, the FNT achieves a time complexity of $\mathcal O(m \overline n |A|)$, where $\overline n$ is the mean of the coordinate-wise maximal polynomial degrees $n_1, \ldots, n_m$ across the $m$ spatial dimensions. In the univariate case, the FNT renders the classic Newton divided difference scheme (DDS). In the multivariate case, however, it improves on the quadratic complexity $\mathcal O(|A|^2)$ of the DDS and on the cost $\mathcal{O}(m \overline n (n_1+1) \cdots (n_m+1))$ of the tensorial interpolation. For sufficiently regular functions, Newton interpolation in Euclidean-degree downward closed polynomial spaces is known to deliver approximation rates equal to those of the tensor product interpolation. Thus, the acceleration power of the FNT comes from requiring substantially fewer degrees of freedom while reaching the same approximation quality as the tensorial interpolation. The inverse transformation has the same time complexity and enables fast evaluation and differentiation of Newton interpolants in quasi-tensorial grids.

math.NA↗

arXiv2025

Second roton feature in the strongly coupled electron liquid

We present extensive \emph{ab initio} path integral Monte Carlo (PIMC) results for the dynamic properties of the finite temperature uniform electron gas (UEG) over a broad range of densities, $2\leq r_s\leq300$. We demonstrate that the direct analysis of the imaginary-time density--density correlation function (ITCF) allows for a rigorous assessment of the density and temperature dependence of the previously reported roton-type feature [T.~Dornheim, \emph{Phys.~Rev.~Lett.}~\textbf{121}, 255001 (2018)] at intermediate wavenumbers. We clearly resolve the emergence of a second roton at the second harmonic of the original feature for $r_s\gtrsim100$, which we identify as an incipient phonon dispersion. Finally, we use our highly accurate PIMC results for the ITCF as the basis for an analytic continuation to compute the dynamic structure factor, which additionally substantiates the existence of the second roton in the strongly coupled electron liquid. Our investigation further elucidates the complex interplay between quantum delocalization and Coulomb coupling in the UEG. All PIMC results are freely available online and provide valuable benchmarks for other theoretical methodologies and approximations.

physics.chem-ph↗

arXiv2025

PyLIT: Reformulation and implementation of the analytic continuation problem using kernel representation methods

Path integral Monte Carlo (PIMC) simulations are a cornerstone for studying quantum many-body systems. The analytic continuation (AC) needed to estimate dynamic quantities from these simulations is an inverse Laplace transform, which is ill-conditioned. If this inversion were surmounted, then dynamical observables (e.g. dynamic structure factor (DSF) $S(q,\omega)$) could be extracted from the imaginary-time correlation functions estimates. Although of important, the AC problem remains challenging due to its ill-posedness. To address this challenge, we express the DSF as a linear combination of kernel functions with known Laplace transforms that have been tailored to satisfy its physical constraints. We use least-squares optimization regularized with a Bayesian prior to determine the coefficients of this linear combination. We explore various regularization term, such as the commonly used entropic regularizer, as well as the Wasserstein distance and $L^2$-distance as well as techniques for setting the regularization weight. A key outcome is the open-source package PyLIT (\textbf{Py}thon \textbf{L}aplace \textbf{I}nverse \textbf{T}ransform), which leverages Numba and unifies the presented formulations. PyLIT's core functionality is kernel construction and optimization. In our applications, we find PyLIT's DSF estimates share qualitative features with other more established methods. We identify three key findings. Firstly, independent of the regularization choice, utilizing non-uniform grid point distributions reduced the number of unknowns and thus reduced our space of possible solutions. Secondly, the Wasserstein distance, a previously unexplored regularizer, performs as good as the entropic regularizer while benefiting from its linear gradient. Thirdly, future work can meaningfully combine regularized and stochastic optimization. (text cut for char. limit)

physics.comp-ph↗

arXiv2025

Multivariate Newton Interpolation in Downward Closed Spaces Reaches the Optimal Geometric Approximation Rates for Bos--Levenberg--Trefethen Functions

We extend the univariate Newton interpolation algorithm to arbitrary spatial dimensions and for any choice of downward-closed polynomial space, while preserving its quadratic runtime and linear storage cost. The generalisation supports any choice of the provided notion of non-tensorial unisolvent interpolation nodes, whose number coincides with the dimension of the chosen-downward closed space. Specifically, we prove that by selecting Leja-ordered Chebyshev-Lobatto or Leja nodes, the optimal geometric approximation rates for a class of analytic functions -- termed Bos--Levenberg--Trefethen functions -- are achieved and extend to the derivatives of the interpolants. In particular, choosing Euclidean degree results in downward-closed spaces whose dimension only grows sub-exponentially with spatial dimension, while delivering approximation rates close to, or even matching those of the tensorial maximum-degree case, mitigating the curse of dimensionality. Several numerical experiments demonstrate the performance of the resulting multivariate Newton interpolation compared to state-of-the-art alternatives and validate our theoretical results.

math.NA↗

arXiv2023

Learning Partial Differential Equations by Spectral Approximates of General Sobolev Spaces

We introduce a novel spectral, finite-dimensional approximation of general Sobolev spaces in terms of Chebyshev polynomials. Based on this polynomial surrogate model (PSM), we realise a variational formulation, solving a vast class of linear and non-linear partial differential equations (PDEs). The PSMs are as flexible as the physics-informed neural nets (PINNs) and provide an alternative for addressing inverse PDE problems, such as PDE-parameter inference. In contrast to PINNs, the PSMs result in a convex optimisation problem for a vast class of PDEs, including all linear ones, in which case the PSM-approximate is efficiently computable due to the exponential convergence rate of the underlying variational gradient descent. As a practical consequence prominent PDE problems were resolved by the PSMs without High Performance Computing (HPC) on a local machine. This gain in efficiency is complemented by an increase of approximation power, outperforming PINN alternatives in both accuracy and runtime. Beyond the empirical evidence we give here, the translation of classic PDE theory in terms of the Sobolev space approximates suggests the PSMs to be universally applicable to well-posed, regular forward and inverse PDE problems.

math.NA↗