arXiv subjects

Stefan Güttel

Publications and source records attributed to Stefan Güttel.

At least 19 recordsLinked to original sources

arXiv2026

ROTE: Benchmarking Neural Memorization on Complexity-Controlled Symbolic Sequences

We introduce ROTE (RollOut Testing of Exact memorization), a benchmarking protocol for evaluating symbolic memorization of neural architectures. We study memorization and the extension of symbolic rules in neural sequence models by using sequences whose complexity is regulated by Lempel--Ziv--Welch (LZW) compression. Under ROTE, each architecture is trained as the same finite-context conditional predictor and is evaluated using teacher-forced one-step prediction as well as closed-loop rollout on the withheld symbols. Following a shared prediction-and-rollout evaluation routine, the benchmark evaluates gated recurrent, minimal recurrent, attention-based, and hybrid recurrent-attention models with their native computational characteristics preserved. Beyond standard predictive metrics, the benchmark reports normalized string distances, training time, memory usage, and parameter count across an LZW-complexity sweep. The study establishes a connection between the complexity of algorithmic sequences and the memorization capacity of neural architectures, revealing the trade-offs involving memorization quality, rollout stability, and computational expense. Our software and reproducible experimental code can be obtained from https://github.com/nla-group/rote.

cs.LG↗

arXiv2026

Integral chemical reaction neural networks

Discovering the structure and kinetics of chemical reaction networks (CRNs) from time-series concentration data is a fundamental challenge in chemical kinetics, with existing approaches relying on prior mechanistic assumptions or suffering from high computational cost and noise sensitivity. In this work, we present integral chemical reaction neural networks (iCRNNs), a framework that combines the interpretable, physics-constrained, architecture of chemical reaction neural networks (CRNNs) with an integral collocation formulation of the governing dynamics. Rather than solving ODEs forward in time at each training step, we approximate the integral of the learned rate functions directly using numerical quadrature. This yields an entirely algebraic forward pass consisting only of matrix operations, eliminating the repeated adaptive ODE solves of standard CRNN and producing smoother, more predictable training. We provide a recovery error analysis characterising the structural sources of ill-conditioning (conservation laws, reaction reversibility, and shared reactant sets) that limit network identifiability. On two benchmark CRNs, iCRNN trains in roughly half the wall-clock time of the baseline CRNN method on a four-species mechanism and between three and four times faster on a five-species mechanism, while completing every training run reliably and attaining comparable or lower loss.

math.NA↗

arXiv2026

Uniform-in-time rational approximation of the matrix exponential with real poles

We propose two new approaches for constructing families of rational functions with shared real poles that nearly uniformly approximate the functions $\exp(-tz)$ for $z\geq 0$ and $t$ in a positive time interval. The first result concerns the case where all real poles coalesce into a single point. With an appropriate choice of a weight function we are able to derive a closed formula for the asymptotically optimal location of such a pole. We then discuss the more general case where all real poles are distinct. Using Zolotarev's construction of certain optimal rational functions, we present a simple algorithm to derive nearly optimal poles efficiently. We analyze the stability of the numerical evaluation of the resulting rational matrix functions in floating-point arithmetic. By controlling the growth of potential ill-conditioning arising from partial fractions, reliable and highly parallelizable exponential propagators are obtained.

math.NA↗

arXiv2026

Sketch-and-Restart: Randomized Sketching in Quadrature-Based Restarting for Matrix Functions

We develop a sketch-and-restart framework for computing the action of a matrix function on a vector, $f(A) b$, where $A$ is large, sparse, and non-Hermitian. The framework combines quadrature-based restarting with Arnoldi-like decompositions generated by sketched or truncated Arnoldi processes. Within this framework, we develop two classes of restarted algorithms. The first uses a fixed Krylov subspace dimension and is based either on the sketched Arnoldi process or on a new sketched harmonic Arnoldi process proposed in this work. The second class chooses the Krylov subspace dimension adaptively by running the truncated Arnoldi process until the condition number of the generated basis, estimated from its sketch, exceeds a prescribed threshold. We also establish the convergence of the restarted sketched harmonic Arnoldi method for Stieltjes functions under the assumption that $A$ is positive real. Numerical experiments demonstrate the effectiveness of the proposed framework, including the computational savings achieved through sketching, the storage reduction enabled by adaptive truncation, and the acceleration obtained from thick restarting.

math.NA↗

arXiv2026

Concentrated real-pole uniform-in-time approximation of the matrix exponential

We propose an asympotically optimal choice of shared concentrated real poles of a family of rational approximants of time-dependent exponential functions $\exp(-tz)$ for $z \geq 0$ and $t$ in a positive time interval $T$. Our result extends a classical result by J.-E. Andersson [J. Approx. Theory, 32(2):85--95, 1981] on the asymptotic best rational approximation of $\exp(-z)$ with real poles. Numerical experiments demonstrate the near-optimality of our choice for various time ranges and for both small and large approximation degrees. An application of the uniform-in-time rational approximation using our proposed concentrated real poles to a linear constant-coefficient initial-value problem is also discussed.

math.NA↗

arXiv2026

Scalable parallel 3-D TEM inversion via rational approximation of the matrix exponential

We present a novel parallel implementation for large-scale three-dimensional electromagnetic inversion based on a Gauss-Newton framework combined with a rational near-best approximation of the matrix exponential for transient simulations. The method employs parallel direct solvers for the shifted linear systems arising from the partial fraction representation of the rational approximation and demonstrates efficient parallel execution on a shared-memory architecture using MPI. A key property of the approach is that the time dependence is entirely contained in the residuals of the employed rational functions, such that the computation of forward responses and sensitivities becomes effectively independent of the number of desired observation times. Model regularization is done with smoothness constraints, formulated with Raviart-Thomas elements. The linearized inverse problems are solved using LSQR, using an implicit parallel Jacobian operator. Numerical experiments demonstrate the successful recovery of a synthetic 3-D conductivity structure with approximately 700,000 degrees of freedom. The study further discusses computational bottlenecks related to memory consumption and shared-memory scalability arising from the simultaneous storage of multiple sparse matrix factorizations. Possible improvements based on preconditioned iterative solvers and distributed high-performance computing architectures are outlined. The implementation in the Julia programming language is released as open-source software to support reproducible research and further development by the geophysical inversion community.

math.NA↗

arXiv2026

Stabilizing randomized GMRES through flexible GMRES

We explore the use of flexible GMRES as an outer wrapper for sketched GMRES. Building on a new bound for the residual of FGMRES in terms of the residual of the preconditioner, we derive a practical randomized solver that requires very little parameter tuning, while still being efficient and robust in the sense of generating non-increasing residual norms.

math.NA↗

arXiv2026

Flexible GMRES converges in two phases

We derive a sharp upper bound on the residuals produced by the flexible GMRES (FGMRES) method. The bound shows that FGMRES exhibits two phases of convergence depending on the residual tolerance of the inner preconditioner. For small tolerances, the convergence of FGMRES is practically geometric with a constant rate throughout, while for looser tolerances the two-phase behavior becomes more pronounced. We also show that the derived bound cannot be improved and construct an example for which it becomes an equality.

math.NA↗

arXiv2026

Data-driven discovery of chemical reaction networks

We propose a unified framework that allows for the full mechanistic reconstruction of chemical reaction networks (CRNs) from concentration data. The framework utilizes an integral formulation of the differential equations governing the chemical reactions, followed by an automatic procedure to recover admissible mass-action mechanisms from the equations. We provide theoretical justification for the use of integral formulations using analytical and numerical error bounds. The integral formulation is demonstrated to offer superior robustness to noise and improved accuracy in both rate-law and graph recovery when compared to other commonly used formulations. Together, our developments advance the goal of fully automated, data-driven chemical mechanism discovery.

math.NA↗

arXiv2026

Fast and explainable clustering in the Manhattan and Tanimoto distance

The CLASSIX algorithm is a fast and explainable approach to data clustering. In its original form, this algorithm exploits the sorting of the data points by their first principal component to truncate the search for nearby data points, with nearness being defined in terms of the Euclidean distance. Here we extend CLASSIX to other distance metrics, including the Manhattan distance and the Tanimoto distance. Instead of principal components, we use an appropriate norm of the data vectors as the sorting criterion, combined with the triangle inequality for search termination. In the case of Tanimoto distance, a provably sharper intersection inequality is used to further boost the performance of the new algorithm. On a real-world chemical fingerprint benchmark, CLASSIX Tanimoto is about 30 times faster than the Taylor--Butina algorithm, and about 80 times faster than DBSCAN, while computing higher-quality clusters in both cases.

cs.LG↗

arXiv2025

Fast parallel transient electromagnetic modeling using a uniform-in-time approximation to the exponential

A new approach for the parallel forward modeling of transient electromagnetic (TEM) fields is presented. It is based on a family of uniform-in-time rational approximants to the matrix exponential that share a common denominator independent of the evaluation time points. The partial fraction decomposition of this family is exploited to devise a fast solver with high parallel efficiency. The number of shifted linear systems that need to be solved in parallel does not depend on the number of required time channels nor the spatial discretization. We also argue that similar parallel efficiency gains can be expected when solving the inverse TEM problem.

math.NA↗

arXiv2024

GMRES with randomized sketching and deflated restarting

We present a new Krylov subspace recycling method for solving a linear system of equations, or a sequence of slowly changing linear systems. Our approach is to reduce the computational overhead of recycling techniques while still benefiting from the acceleration afforded by such techniques. As such, this method augments an unprojected Krylov subspace. Furthermore, it combines randomized sketching and deflated restarting in a way that avoids orthogononalizing a full Krylov basis. We call this new method GMRES-SDR (sketched deflated restarting). With this new method, we provide new theory, which initially characterizes unaugmented sketched GMRES as a projection method for which the projectors involve the sketching operator. We demonstrate that sketched GMRES and its sibling method sketched FOM are an MR/OR pairing, just like GMRES and FOM. We furthermore obtain residual convergence estimates. Building on this, we characterize GMRES-SDR also in terms of sketching-based projectors. Compression of the augmented Krylov subspace for recycling is performed using a sketched version of harmonic Ritz vectors. We present results of numerical experiments demonstrating the effectiveness of GMRES-SDR over competitor methods such as GMRES-DR and GCRO-DR.

math.NA↗

arXiv2024

A Sherman--Morrison--Woodbury approach to solving least squares problems with low-rank updates

We present a simple formula to update the pseudoinverse of a full-rank rectangular matrix that undergoes a low-rank modification, and demonstrate its utility for solving least squares problems. The resulting algorithm can be dramatically faster than solving the modified least squares problem from scratch, just like the speedup enabled by Sherman--Morrison--Woodbury for solving linear systems with low-rank modifications.

math.NA↗

arXiv2024

Randomized sketching of nonlinear eigenvalue problems

Rational approximation is a powerful tool to obtain accurate surrogates for nonlinear functions that are easy to evaluate and linearize. The interpolatory adaptive Antoulas--Anderson (AAA) method is one approach to construct such approximants numerically. For large-scale vector- and matrix-valued functions, however, the direct application of the set-valued variant of AAA becomes inefficient. We propose and analyze a new sketching approach for such functions called sketchAAA that, with high probability, leads to much better approximants than previously suggested approaches while retaining efficiency. The sketching approach works in a black-box fashion where only evaluations of the nonlinear function at sampling points are needed. Numerical tests with nonlinear eigenvalue problems illustrate the efficacy of our approach, with speedups above 200 for sampling large-scale black-box functions without sacrificing on accuracy.

math.NA↗

arXiv2024

A sketch-and-select Arnoldi process

A sketch-and-select Arnoldi process to generate a well-conditioned basis of a Krylov space at low cost is proposed. At each iteration the procedure utilizes randomized sketching to select a limited number of previously computed basis vectors to project out of the current basis vector. The computational cost grows linearly with the dimension of the Krylov space. The subset selection problem for the projection step is approximately solved with a number of heuristic algorithms and greedy methods used in statistical learning and compressive sensing.

math.NA↗

arXiv2024

Fast and explainable clustering based on sorting

We introduce a fast and explainable clustering method called CLASSIX. It consists of two phases, namely a greedy aggregation phase of the sorted data into groups of nearby data points, followed by the merging of groups into clusters. The algorithm is controlled by two scalar parameters, namely a distance parameter for the aggregation and another parameter controlling the minimal cluster size. Extensive experiments are conducted to give a comprehensive evaluation of the clustering performance on synthetic and real-world datasets, with various cluster shapes and low to high feature dimensionality. Our experiments demonstrate that CLASSIX competes with state-of-the-art clustering algorithms. The algorithm has linear space complexity and achieves near linear time complexity on a wide range of problems. Its inherent simplicity allows for the generation of intuitive explanations of the computed clusters.

cs.LG↗

arXiv2024

Fast and exact fixed-radius neighbor search based on sorting

Fixed-radius near neighbor search is a fundamental data operation that retrieves all data points within a user-specified distance to a query point. There are efficient algorithms that can provide fast approximate query responses, but they often have a very compute-intensive indexing phase and require careful parameter tuning. Therefore, exact brute force and tree-based search methods are still widely used. Here we propose a new fixed-radius near neighbor search method, called SNN, that significantly improves over brute force and tree-based methods in terms of index and query time, provably returns exact results, and requires no parameter tuning. SNN exploits a sorting of the data points by their first principal component to prune the query search space. Further speedup is gained from an efficient implementation using high-level Basic Linear Algebra Subprograms (BLAS). We provide theoretical analysis of our method and demonstrate its practical performance when used stand-alone and when applied within the DBSCAN clustering algorithm.

cs.IR↗

arXiv2023

Krylov Subspace Recycling With Randomized Sketching For Matrix Functions

A Krylov subspace recycling method for the efficient evaluation of a sequence of matrix functions acting on a set of vectors is developed. The method improves over the recycling methods presented in [Burke et al., arXiv:2209.14163, 2022] in that it uses a closed-form expression for the augmented FOM approximants and hence circumvents the use of numerical quadrature. We further extend our method to use randomized sketching in order to avoid the arithmetic cost of orthogonalizing a full Krylov basis, offering an attractive solution to the fact that recycling algorithms built from shifted augmented FOM cannot easily be restarted. The efficacy of the proposed algorithms is demonstrated with numerical experiments.

math.NA↗