Search

We introduce a method that combines neural operators, physics-informed machine learning, and standard numerical methods for solving PDEs. The proposed approach extends each of the aforementioned methods and unifies them within a single framework. We can parametrically solve partial differential equations in a data-free manner and provide accurate sensitivities, meaning the derivatives of the solution space with respect to the design space. These capabilities enable gradient-based optimization without the typical sensitivity analysis costs, unlike adjoint methods that scale directly with the number of response functions. Our Finite Operator Learning (FOL) approach uses an uncomplicated feed-forward neural network model to directly map the discrete design space (i.e. parametric input space) to the discrete solution space (i.e. finite number of sensor points in the arbitrary shape domain) ensuring compliance with physical laws by designing them into loss functions. The discretized governing equations, as well as the design and solution spaces, can be derived from any well-established numerical techniques. In this work, we employ the Finite Element Method (FEM) to approximate fields and their spatial derivatives. Subsequently, we conduct Sobolev training to minimize a multi-objective loss function, which includes the discretized weak form of the energy functional, boundary conditions violations, and the stationarity of the residuals with respect to the design variables. Our study focuses on the steady-state heat equation within heterogeneous materials that exhibits significant phase contrast and possibly temperature-dependent conductivity. The network's tangent matrix is directly used for gradient-based optimization to improve the microstructure's heat transfer characteristics. ..., Comment: arXiv admin note: text overlap with arXiv:2401.02363

Modeling structure and behavior of software systems plays a crucial role in the industrial practice of software engineering. As with other software engineering artifacts, software models are subject to evolution. Supporting modelers in evolving software models with recommendations for model completions is still an open problem, though. In this paper, we explore the potential of large language models for this task. In particular, we propose an approach, retrieval-augmented generation, leveraging large language models, model histories, and retrieval-augmented generation for model completion. Through experiments on three datasets, including an industrial application, one public open-source community dataset, and one controlled collection of simulated model repositories, we evaluate the potential of large language models for model completion with retrieval-augmented generation. We found that large language models are indeed a promising technology for supporting software model evolution (62.30% semantically correct completions on real-world industrial data and up to 86.19% type-correct completions). The general inference capabilities of large language models are particularly useful when dealing with concepts for which there are few, noisy, or no examples at all.

We propose a novel finite element-based physics-informed operator learning framework that allows for predicting spatiotemporal dynamics governed by partial differential equations (PDEs). The proposed framework employs a loss function inspired by the finite element method (FEM) with the implicit Euler time integration scheme. A transient thermal conduction problem is considered to benchmark the performance. 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. The Galerkin discretized weak formulation of the heat equation is employed to incorporate physics into the loss function, which is coined finite operator learning (FOL). Upon training, the networks successfully predict the temperature evolution over time for any initial temperature field at high accuracy compared to the FEM solution. The framework is also confirmed to be applicable to a heterogeneous thermal conductivity and arbitrary geometry. The advantages of FOL can be summarized as follows: First, the training is performed in an unsupervised manner, avoiding the need for a large data set prepared from costly simulations or experiments. Instead, random temperature patterns generated by the Gaussian random process and the Fourier series, combined with constant temperature fields, are used as training data to cover possible temperature cases. Second, shape functions and backward difference approximation are exploited for the domain discretization, resulting in a purely algebraic equation. This enhances training efficiency, as one avoids time-consuming automatic differentiation when optimizing weights and biases while accepting possible discretization errors. Finally, thanks to the interpolation power of FEM, any arbitrary geometry can be handled with FOL, which is crucial to addressing various engineering application scenarios.

To obtain fast solutions for governing physical equations in solid mechanics, we introduce a method that integrates the core ideas of the finite element method with physics-informed neural networks and concept of neural operators. This approach generalizes and enhances each method, learning the parametric solution for mechanical problems without relying on data from other resources (e.g. other numerical solvers). We propose directly utilizing the available discretized weak form in finite element packages to construct the loss functions algebraically, thereby demonstrating the ability to find solutions even in the presence of sharp discontinuities. Our focus is on micromechanics as an example, where knowledge of deformation and stress fields for a given heterogeneous microstructure is crucial for further design applications. The primary parameter under investigation is the Young's modulus distribution within the heterogeneous solid system. Our investigations reveal that physics-based training yields higher accuracy compared to purely data-driven approaches for unseen microstructures. Additionally, we offer two methods to directly improve the process of obtaining high-resolution solutions, avoiding the need to use basic interpolation techniques. First is based on an autoencoder approach to enhance the efficiency for calculation on high resolution grid point. Next, Fourier-based parametrization is utilized to address complex 2D and 3D problems in micromechanics. The latter idea aims to represent complex microstructures efficiently using Fourier coefficients. Comparisons with other well-known operator learning algorithms, further emphasize the advantages of the newly proposed method.

The article examines a linear-quadratic Neumann control problem that is governed by a non-coercive elliptic equation. Due to the non-self-adjoint nature of the linear control-to-state operator, it is necessary to independently study both the state and adjoint state equations. The article establishes the existence and uniqueness of solutions for both equations, with minimal assumptions made about the problem's data. Next, the regularity of these solutions is studied in three frameworks: Hilbert-Sobolev spaces, Sobolev-Slobodecki\u\i{} spaces, and weighted Sobolev spaces. These regularity results enable a numerical analysis of the finite element approximation of both the state and adjoint state equations. The results cover both convex and non-convex domains and quasi-uniform and graded meshes. Finally, the optimal control problem is analyzed and discretized. Existence and uniqueness of the solution, first-order optimality conditions, and error estimates for the finite element approximation of the control are obtained. Numerical experiments confirming these results are included. A significant highlight is that the discretization error estimates known from the literature, are improved even for the coercive case.

The Laplace eigenvalue problem on circular sectors has eigenfunctions with corner singularities. Standard methods may produce suboptimal approximation results. To address this issue, a novel numerical algorithm that enhances standard isogeometric analysis is proposed in this paper by using a single-patch graded mesh refinement scheme. Numerical tests demonstrate optimal convergence rates for both the eigenvalues and eigenfunctions. Furthermore, the results show that smooth splines possess a superior approximation constant compared to their $C^0$-continuous counterparts for the lower part of the Laplace spectrum. This is an extension of previous findings about excellent spectral approximation properties of smooth splines on rectangular domains to circular sectors. In addition, graded meshes prove to be particularly advantageous for an accurate approximation of a limited number of eigenvalues. The novel algorithm applied here has a drawback in the singularity of the isogeometric parameterization. It results in some basis functions not belonging to the solution space of the corresponding weak problem, which is considered a variational crime. However, the approach proves to be robust. Finally, a hierarchical mesh structure is presented to avoid anisotropic elements, omit redundant degrees of freedom and keep the number of basis functions contributing to the variational crime constant, independent of the mesh size. Numerical results validate the effectiveness of hierarchical mesh grading for the simulation of eigenfunctions with and without corner singularities., Comment: 35 pages, 19 figures

Valid empirical studies build confidence in scientific findings. Fortunately, it is now common for software engineering researchers to consider threats to validity when designing their studies and to discuss them as part of their publication. Yet, in complex experiments with human participants, there is often an overwhelming number of intuitively plausible threats to validity -- more than a researcher can feasibly cover. Therefore, prioritizing potential threats to validity becomes crucial. We suggest moving away from relying solely on intuition for prioritizing validity threats, and propose that evidence on the actual impact of suspected threats to validity should complement intuition., Comment: Accepted to WSESE 2024, an ICSE co-located workshop on methodological issues with empirical studies in software engineering

Lipolysis is a life-essential metabolic process, which supplies fatty acids stored in lipid droplets to the body in order to match the demands of building new cells and providing cellular energy. In this paper, we present a first mathematical modelling approach for lipolysis, which takes into account that the involved enzymes act on the surface of lipid droplets. We postulate an active region near the surface where the substrates are within reach of the surface-bound enzymes and formulate a system of reaction-diffusion PDEs, which connect the active region to the inner core of lipid droplets via interface conditions. We establish two numerical discretisations based on finite element method and isogeometric analysis, and validate them to perform reliably. For numerical testing purposes, we introduce and analyse a testing model featuring a nontrivial explicit stationary state solution, which describes beside lipolysis also a reverse process (in a physiologically oversimplified way). We prove the unique existence of global and equilibrium solutions. We establish exponential convergence to the equilibrium solutions using the entropy method. We then study the stationary state model and compute explicitly for radially symmetric solutions. Concerning the finite element methods, we show numerically the linear and quadratic convergence of the errors with respect to the $H^{1}$- and $L^{2}$-norms, respectively. Finally, we present numerical simulations of a prototypical PDE model of lipolysis and illustrate that ATGL clustering on lipid droplets can significantly slow down lipolysis., Comment: 27 pages, 18 figures

We investigate the link between quantum position-verification (QPV) and holography established in [MPS19] using holographic quantum error correcting codes as toy models. By inserting the "temporal" scaling of the AdS metric by hand via the bulk Hamiltonian interaction strength, we recover a toy model with consistent causality structure. This leads to an interesting implication between two topics in quantum information: if position-based verification is secure against attacks with small entanglement then there are new fundamental lower bounds for resources required for one Hamiltonian to simulate another., Comment: (v2: updated presentation)

Background: In times when the ability to program is becoming increasingly important, it is still difficult to teach students to become successful programmers. One remarkable aspect are recent findings from neuro-imaging studies, which suggest a consistent role of language competency of novice programmers when they learn programming. Thus, for effectively teaching programming, it might be beneficial to draw from linguistic research, especially from foreign language acquisition. Objective: The goal of this study is to investigate the feasibility of this idea, such that we can enhance learning programming by activating language learning mechanisms. Method: To this end, we conducted an empirical study, in which we taught one group of students an artificial language, while another group received an introduction into Git as control condition, before we taught both groups basic programming knowledge in a programming course. Result: We observed that the training of the artificial language can be easily integrated into our curriculum. Furthermore, we observed that language learning strategies were activated and that participants perceived similarities between learning the artificial language and the programming language. However, within the context of our study, we did not find a significant benefit for programming competency when students learned an artificial language first. Conclusion: Our study lays the methodological foundation to explore the use of natural language acquisition research and expand this field step by step. We report our experience here to guide research and to open up the possibilities from the field of linguistic research to improve programming acquisition., Comment: 13 pages, 5 figures

We present a method that employs physics-informed deep learning techniques for parametrically solving partial differential equations. The focus is on the steady-state heat equations within heterogeneous solids exhibiting significant phase contrast. Similar equations manifest in diverse applications like chemical diffusion, electrostatics, and Darcy flow. The neural network aims to establish the link between the complex thermal conductivity profiles and temperature distributions, as well as heat flux components within the microstructure, under fixed boundary conditions. A distinctive aspect is our independence from classical solvers like finite element methods for data. A noteworthy contribution lies in our novel approach to defining the loss function, based on the discretized weak form of the governing equation. This not only reduces the required order of derivatives but also eliminates the need for automatic differentiation in the construction of loss terms, accepting potential numerical errors from the chosen discretization method. As a result, the loss function in this work is an algebraic equation that significantly enhances training efficiency. We benchmark our methodology against the standard finite element method, demonstrating accurate yet faster predictions using the trained neural network for temperature and flux profiles. We also show higher accuracy by using the proposed method compared to purely data-driven approaches for unforeseen scenarios.

Modeling structure and behavior of software systems plays a crucial role, in various areas of software engineering. As with other software engineering artifacts, software models are subject to evolution. Supporting modelers in evolving models by model completion facilities and providing high-level edit operations such as frequently occurring editing patterns is still an open problem. Recently, large language models (i.e., generative neural networks) have garnered significant attention in various research areas, including software engineering. In this paper, we explore the potential of large language models in supporting the evolution of software models in software engineering. We propose an approach that utilizes large language models for model completion and discovering editing patterns in model histories of software systems. Through controlled experiments using simulated model repositories, we conduct an evaluation of the potential of large language models for these two tasks. We have found that large language models are indeed a promising technology for supporting software model evolution, and that it is worth investigating further in the area of software model evolution.

While LDPC codes have been demonstrated with desirable error correcting properties, this has come at a cost of diverging from the geometrical constraints of many hardware platforms. Viewing codes as the groundspace of a Hamiltonian, we consider engineering a simulation Hamiltonian reproducing some relevant features of the code. Techniques from Hamiltonian simulation theory are used to build a simulation of LDPC codes using only 2D nearest-neighbour interactions at the cost of an energy penalty polynomial in the system size. We derive guarantees for the simulation that allows us to approximately reproduce the ground state of the code Hamiltonian, approximating a $[[N, \Omega(\sqrt{N}), \Omega(\sqrt{N})]]$ code in 2D. The key ingredient is a new constructive tool to simulate an $l$-long interaction between two qubits by a 1D chain of $l$ nearest-neighbour interacting qubits using $\mathrm{poly}( l)$ interaction strengths. This is an exponential advantage over the existing gadgets for this routine which facilitates the first $\epsilon$-simulation of \emph{arbitrary sparse} Hamiltonian on $n$ qubits with a Hamiltonian on a 2D lattice of $O(n^2)$ qubits with interaction strengths scaling as $O\left(\mathrm{poly}(n,1/\epsilon)\right)$.

Recent advances in Large Language Models (LLMs) have spurred interest in designing LLM-based agents for tasks that involve interaction with human and artificial agents. This paper addresses a key aspect in the design of such agents: Predicting human decision in off-policy evaluation (OPE), focusing on language-based persuasion games, where the agent's goal is to influence its partner's decisions through verbal messages. Using a dedicated application, we collected a dataset of 87K decisions from humans playing a repeated decision-making game with artificial agents. Our approach involves training a model on human interactions with one agents subset to predict decisions when interacting with another. To enhance off-policy performance, we propose a simulation technique involving interactions across the entire agent space and simulated decision makers. Our learning strategy yields significant OPE gains, e.g., improving prediction accuracy in the top 15% challenging cases by 7.1%. Our code and the large dataset we collected and generated are submitted as supplementary material and publicly available in our GitHub repository: https://github.com/eilamshapira/HumanChoicePrediction

We present MeeQA, a dataset for natural-language question answering over meeting transcripts. It includes real questions asked during meetings by its participants. The dataset contains 48K question-answer pairs, extracted from 422 meeting transcripts, spanning multiple domains. Questions in transcripts pose a special challenge as they are not always clear, and considerable context may be required in order to provide an answer. Further, many questions asked during meetings are left unanswered. To improve baseline model performance on this type of questions, we also propose a novel loss function, \emph{Flat Hierarchical Loss}, designed to enhance performance over questions with no answer in the text. Our experiments demonstrate the advantage of using our approach over standard QA models.

Background: Despite similar education and background, programmers can exhibit vast differences in efficacy. While research has identified some potential factors, such as programming experience and domain knowledge, the effect of these factors on programmers' efficacy is not well understood. Aims: We aim at unraveling the relationship between efficacy (speed and correctness) and measures of programming experience. We further investigate the correlates of programmer efficacy in terms of reading behavior and cognitive load. Method: For this purpose, we conducted a controlled experiment with 37~participants using electroencephalography (EEG) and eye tracking. We asked participants to comprehend up to 32 Java source-code snippets and observed their eye gaze and neural correlates of cognitive load. We analyzed the correlation of participants' efficacy with popular programming experience measures. Results: We found that programmers with high efficacy read source code more targeted and with lower cognitive load. Commonly used experience levels do not predict programmer efficacy well, but self-estimation and indicators of learning eagerness are fairly accurate. Implications: The identified correlates of programmer efficacy can be used for future research and practice (e.g., hiring). Future research should also consider efficacy as a group sampling method, rather than using simple experience measures., Comment: 12 pages, 7 figures, To be published at Joint European Software Engineering Conference and Symposium on the Foundations of Software Engineering (ESEC/FSE '22)

We investigate a variant of the SSH model consisting of an SSH chain with an embedded Aharonov-Bohm quantum ring. The embedded ring gives rise to domain wall states whose energy levels are in the band gap. The dependence of some of the states on the ring's magnetic flux provides a mechanism to control the entanglement between the two SSH edge fermion modes. The concurrence between the edge modes depends periodically on the magnetic flux. In the deep topological zone, for appropriate values of the flux, the concurrence reaches the extreme values of $0$ and $1$, corresponding to vanishing entanglement and to maximum entanglement. Therefore, the QR-SSH system offers a tuning mechanism that can be used either as an entanglement modulator or as an entanglement switch., Comment: 13 pages 6 figures

The LOPES experiment was a radio interferometer built at the existing air shower array KASCADE-Grande in Karlsruhe, Germany. The last configuration of LOPES was called LOPES 3D and consisted of ten tripole antennas. Each of these antennas consisted of three crossed dipoles east-west, north-south, and vertically aligned. With this, LOPES 3D had the unique possibility to study the benefits of measurements with vertically aligned antennas in the environment of the well understood and calibrated particle detector array KASCADE-Grande. The measurements with three spatially coincident antennas allows a redundant reconstruction of the electric field vector. Several methods to exploit the redundancy were developed and tested. Furthermore, for the first time in LOPES, the background noise could be studied polarization- and direction dependent. With LOPES 3D it could be demonstrated that radio detection reaches a higher efficiency for inclined showers when including measurements with vertically aligned antennas and that the vertical component gets more important for the measurement of inclined showers. In this contribution we discuss a weighting scheme for the best combination of three redundant reconstructed electric field vectors. Furthermore, we discuss the influence of these weighting schemes on the ability to reconstruct air showers using the radio method. We show an estimate of the radio efficiency for inclined showers with focus on the benefits of measurements with vertically aligned antennas and we present the direction dependent noise in the different polarizations., Comment: Submission to the proceedings of the ARENA2014 workshop, which unfortunately were never published; thus documented here for reference

Analogue Hamiltonian simulation is a promising near-term application of quantum computing and has recently been put on a theoretical footing. In Hamiltonian simulation, a physical Hamiltonian is engineered to have identical physics to another - often very different - Hamiltonian. This is qualitatively similar to the notion of duality in physics, whereby two superficially different theories are mathematically equivalent in some precise sense. However, existing characterisations of Hamiltonian simulations are not sufficiently general to extend to all dualities in physics. In particular, they cannot encompass the important cases of strong/weak and high-temperature/low-temperature dualities. In this work, we give three physically motivated axiomatisations of duality, formulated respectively in terms of observables, partition functions and entropies. We prove that these axiomatisations are equivalent, and characterise the mathematical form that any duality satisfying these axioms must take. A building block in one of our results is a strengthening of earlier results on entropy-preserving maps to maps that are entropy-preserving up to an additive constant, which we prove decompose as a direct sum of unitary and anti-unitary components, which may be of independent mathematical interest., Comment: 40 pages. v3: after peer review modifications