Search

We review recent and present new results on thermalization of nonabelian gauge theory obtained by exact numerical simulation of the real-time dynamics of two-dimensional SU(2) lattice gauge theory. We discuss: (1) tests confirming the Eigenstate Thermalization Hypothesis; (2) the entanglement entropy of sublattices, including the Page curve, the transition from area to volume scaling with increasing energy of the eigenstate and its time evolution that shows thermalization of localized regions to be a two-step process; (3) the absence of quantum many-body scars for $j_{\rm max}>1/2$; (4) the spectral form factor, which exhibits the expected slope-ramp-plateau structure for late times; (5) the entanglement Hamiltonian for SU(2), which has properties in accordance with the Bisognano-Wichmann theorem; and (6) a measure for non-stabilizerness or "magic" that is found to reach its maximum during thermalization. We conclude that the thermalization of nonabelian gauge theories is a promising process to establish quantum advantage., Comment: Latex error corrected

In-basket message interactions play a crucial role in physician-patient communication, occurring during all phases (pre-, during, and post) of a patient's care journey. However, responding to these patients' inquiries has become a significant burden on healthcare workflows, consuming considerable time for clinical care teams. To address this, we introduce RadOnc-GPT, a specialized Large Language Model (LLM) powered by GPT-4 that has been designed with a focus on radiotherapeutic treatment of prostate cancer with advanced prompt engineering, and specifically designed to assist in generating responses. We integrated RadOnc-GPT with patient electronic health records (EHR) from both the hospital-wide EHR database and an internal, radiation-oncology-specific database. RadOnc-GPT was evaluated on 158 previously recorded in-basket message interactions. Quantitative natural language processing (NLP) analysis and two grading studies with clinicians and nurses were used to assess RadOnc-GPT's responses. Our findings indicate that RadOnc-GPT slightly outperformed the clinical care team in "Clarity" and "Empathy," while achieving comparable scores in "Completeness" and "Correctness." RadOnc-GPT is estimated to save 5.2 minutes per message for nurses and 2.4 minutes for clinicians, from reading the inquiry to sending the response. Employing RadOnc-GPT for in-basket message draft generation has the potential to alleviate the workload of clinical care teams and reduce healthcare costs by producing high-quality, timely responses.

This paper presents a general framework integrating vision and acoustic sensor data to enhance localization and mapping in highly dynamic and complex underwater environments, with a particular focus on fish farming. The proposed pipeline is suited to obtain both the net-relative pose estimates of an Unmanned Underwater Vehicle (UUV) and the depth map of the net pen purely based on vision data. Furthermore, this paper presents a method to estimate the global pose of an UUV fusing the net-relative pose estimates with acoustic data. The pipeline proposed in this paper showcases results on datasets obtained from industrial-scale fish farms and successfully demonstrates that the vision-based TRU-Depth model, when provided with sparse depth priors from the FFT method and combined with the Wavemap method, can estimate both net-relative and global position of the UUV in real time and generate detailed 3D maps suitable for autonomous navigation and inspection purposes.

Building-specific knowledge such as building type and function information is important for numerous energy applications. However, comprehensive datasets containing this information for individual households are missing in many regions of Europe. For the first time, we investigate whether it is feasible to predict building types and functional classes at a European scale based on only open GIS datasets available across countries. We train a graph neural network (GNN) classifier on a large-scale graph dataset consisting of OpenStreetMap (OSM) buildings across the EU, Norway, Switzerland, and the UK. To efficiently perform training using the large-scale graph, we utilize localized subgraphs. A graph transformer model achieves a high Cohen's kappa coefficient of 0.754 when classifying buildings into 9 classes, and a very high Cohen's kappa coefficient of 0.844 when classifying buildings into the residential and non-residential classes. The experimental results imply three core novel contributions to literature. Firstly, we show that building classification across multiple countries is possible using a multi-source dataset consisting of information about 2D building shape, land use, degree of urbanization, and countries as input, and OSM tags as ground truth. Secondly, our results indicate that GNN models that consider contextual information about building neighborhoods improve predictive performance compared to models that only consider individual buildings and ignore the neighborhood. Thirdly, we show that training with GNNs on localized subgraphs instead of standard GNNs improves performance for the task of building classification.

Nonabelian gauge theories are chaotic in the classical limit. We discuss new evidence from SU(2) lattice gauge theory that they are also chaotic at the quantum level. We also describe possible future studies aimed at discovering the consequences of this insight., Comment: Lecture presented at the Particles and Plasmas Symposium 2024, Budapest (Hungary)

Finding the eigenvalues connected to the covariance operator of a centred Hilbert-space valued Gaussian process is genuinely considered a hard problem in several mathematical disciplines. In statistics this problem arises for instance in the asymptotic null distribution of goodness-of-fit test statistics of weighted $L^2$-type. For this problem we present the Rayleigh-Ritz method to approximate the eigenvalues. The usefulness of these approximations is shown by high lightening implications such as critical value approximation and theoretical comparison of test statistics by means of Bahadur efficiencies., Comment: 18 pages, 16 tables

To enhance the spatial resolution of the M\"ONCH 25 \textmu m pitch hybrid pixel detector, deep learning models have been trained using both simulation and measurement data. Challenges arise when comparing simulation-based deep learning models to measurement-based models for electrons, as the spatial resolution achieved through simulations is notably inferior to that from measurements. Discrepancies are also observed when directly comparing X-ray simulations with measurements, particularly in the spectral output of single pixels. These observations collectively suggest that current simulations require optimization. To address this, the dynamics of charge carriers within the silicon sensor have been studied using Monte Carlo simulations, aiming to refine the charge transport modeling. The simulation encompasses the initial generation of the charge cloud, charge cloud drift, charge diffusion and repulsion, and electronic noise. The simulation results were validated with measurements from the M\"ONCH detector for X-rays, and the agreement between measurements and simulations was significantly improved by accounting for the charge repulsion., Comment: Prepared for submission to JINST as a proceeding for 25th International Workshops on Radiation Imaging Detectors

Hallucinations pose a challenge to the application of large language models (LLMs) thereby motivating the development of metrics to evaluate factual precision. We observe that popular metrics using the Decompose-Then-Verify framework, such as \FActScore, can be manipulated by adding obvious or repetitive subclaims to artificially inflate scores. This observation motivates our new customizable plug-and-play subclaim selection component called Core, which filters down individual subclaims according to their uniqueness and informativeness. We show that many popular factual precision metrics augmented by Core are substantially more robust on a wide range of knowledge domains. We release an evaluation framework supporting easy and modular use of Core and various decomposition strategies, which we recommend adoption by the community. We also release an expansion of the FActScore biography dataset to facilitate further studies of decomposition-based factual precision evaluation.

Inverse problems are fundamental in fields like medical imaging, geophysics, and computerized tomography, aiming to recover unknown quantities from observed data. However, these problems often lack stability due to noise and ill-conditioning, leading to inaccurate reconstructions. To mitigate these issues, regularization methods are employed, introducing constraints to stabilize the inversion process and achieve a meaningful solution. Recent research has shown that the application of regularizing filters to diagonal frame decompositions (DFD) yields regularization methods. These filters dampen some frame coefficients to prevent noise amplification. This paper introduces a non-linear filtered DFD method combined with a learning strategy for determining optimal non-linear filters from training data pairs. In our experiments, we applied this approach to the inversion of the Radon transform using 500 image-sinogram pairs from real CT scans. Although the learned filters were found to be strictly increasing, they did not satisfy the non-expansiveness condition required to link them with convex regularizers and prove stability and convergence in the sense of regularization methods in previous works. Inspired by this, the paper relaxes the non-expansiveness condition, resulting in weakly convex regularization. Despite this relaxation, we managed to derive stability, convergence, and convergence rates with respect to the absolute symmetric Bregman distance for the learned non-linear regularizing filters. Extensive numerical results demonstrate the effectiveness of the proposed method in achieving stable and accurate reconstructions., Comment: The full Python code for learning the weakly convex regularizing filters will be publicly available at https://github.com/matthi99/learned-filters.git

Proof-oriented programs mix computational content with proofs of program correctness. However, the human effort involved in programming and proving is still substantial, despite the use of Satisfiability Modulo Theories (SMT) solvers to automate proofs in languages such as F*. Seeking to spur research on using AI to automate the construction of proof-oriented programs, we curate a dataset of 600K lines of open-source F* programs and proofs, including software used in production systems ranging from Windows and Linux to Python and Firefox. Our dataset includes around 32K top-level F* definitions, each representing a type-directed program and proof synthesis problem producing a definition given a formal specification expressed as an F* type. We provide a program fragment checker that queries F* to check the correctness of candidate solutions. We also report on an extended version of our dataset containing a total of 940K lines of programs and proofs, with a total of 54k top-level F* definitions. We believe this is the largest corpus of SMT-assisted program proofs coupled with a reproducible program-fragment checker. Grounded in this dataset, we investigate the use of AI to synthesize programs and their proofs in F*, with promising results. Our main finding in that the performance of fine-tuned smaller language models (such as Phi-2 or StarCoder) compare favorably with large language models (such as GPT-4), at a much lower computational cost. We also identify various type-based retrieval augmentation techniques and find that they boost performance significantly. With detailed error analysis and case studies, we identify potential strengths and weaknesses of models and techniques and suggest directions for future improvements., Comment: 47th International Conference on Software Engineering

As generated text becomes more commonplace, it is increasingly important to evaluate how well-supported such text is by external knowledge sources. Many approaches for evaluating textual support rely on some method for decomposing text into its individual subclaims which are scored against a trusted reference. We investigate how various methods of claim decomposition -- especially LLM-based methods -- affect the result of an evaluation approach such as the recently proposed FActScore, finding that it is sensitive to the decomposition method used. This sensitivity arises because such metrics attribute overall textual support to the model that generated the text even though error can also come from the metric's decomposition step. To measure decomposition quality, we introduce an adaptation of FActScore, which we call DecompScore. We then propose an LLM-based approach to generating decompositions inspired by Bertrand Russell's theory of logical atomism and neo-Davidsonian semantics and demonstrate its improved decomposition quality over previous methods.

We study the entanglement entropy of Hamiltonian SU(2) lattice gauge theory in $2+1$ dimensions on linear plaquette chains and show that the entanglement entropies of both ground and excited states follow Page curves. The transition of the subsystem size dependence of the entanglement entropy from the area law for the ground state to the volume law for highly excited states is found to be described by a universal crossover function. Quantum many-body scars in the middle of the spectrum, which are present in the electric flux truncated Hilbert space, where the gauge theory can be mapped onto an Ising model, disappear when higher electric field representations are included in the Hilbert space basis. This suggests the continuum $(2+1)$-dimensional SU(2) gauge theory does not have such scarred states., Comment: 14 pages, 16 figures; v2: update the discussions of area-volume transition and scars; v3: published version; v4: corrected a typo in the legend of Fig.7

Spherical and hyperspherical data are commonly encountered in diverse applied research domains, underscoring the vital task of assessing independence within such data structures. In this context, we investigate the properties of test statistics relying on distance correlation measures originally introduced for the energy distance, and generalize the concept to strongly negative definite kernel-based distances. An important benefit of employing this method lies in its versatility across diverse forms of directional data, enabling the examination of independence among vectors of varying types. The applicability of tests is demonstrated on several real datasets.

We report on progress in full quantum understanding of thermalization in non-Abelian gauge theories. Specifically, we test the eigenstate thermalization hypothesis for (2+1)-dimensional SU(2) lattice gauge theory., Comment: 4 pages, 4 figures, contribution to the proceedings of the 30th International Conference on Ultra-relativistic Nucleus-Nucleus Collisions (Quark Matter 2023)

Color constancy and color illusion perception are two phenomena occurring in the human visual system, which can help us reveal unknown mechanisms of human perception. For decades computer vision scientists have developed numerous color constancy methods, which estimate the reflectance of the surface by discounting the illuminant. However, color illusions have not been analyzed in detail in the field of computational color constancy, which we find surprising since the relationship they share is significant and may let us design more robust systems. We argue that any model that can reproduce our sensation on color illusions should also be able to provide pixel-wise estimates of the light source. In other words, we suggest that the analysis of color illusions helps us to improve the performance of the existing global color constancy methods, and enable them to provide pixel-wise estimates for scenes illuminated by multiple light sources. In this study, we share the outcomes of our investigation in which we take several color constancy methods and modify them to reproduce the behavior of the human visual system on color illusions. Also, we show that parameters purely extracted from illusions are able to improve the performance of color constancy methods. A noteworthy outcome is that our strategy based on the investigation of color illusions outperforms the state-of-the-art methods that are specifically designed to transform global color constancy algorithms into multi-illuminant algorithms., Comment: This work is accepted at VISAPP 2024 as a long paper

In this work, we address the problem of real-time dense depth estimation from monocular images for mobile underwater vehicles. We formulate a deep learning model that fuses sparse depth measurements from triangulated features to improve the depth predictions and solve the problem of scale ambiguity. To allow prior inputs of arbitrary sparsity, we apply a dense parameterization method. Our model extends recent state-of-the-art approaches to monocular image based depth estimation, using an efficient encoder-decoder backbone and modern lightweight transformer optimization stage to encode global context. The network is trained in a supervised fashion on the forward-looking underwater dataset, FLSea. Evaluation results on this dataset demonstrate significant improvement in depth prediction accuracy by the fusion of the sparse feature priors. In addition, without any retraining, our method achieves similar depth prediction accuracy on a downward looking dataset we collected with a diver operated camera rig, conducting a survey of a coral reef. The method achieves real-time performance, running at 160 FPS on a laptop GPU and 7 FPS on a single CPU core and is suitable for direct deployment on embedded systems. The implementation of this work is made publicly available at https://github.com/ebnerluca/uw_depth., Comment: Submitted to ICRA 2024

Experiments at synchrotron radiation sources and X-ray Free-Electron Lasers in the soft X-ray energy range ($250$eV--$2$keV) stand to benefit from the adaptation of the hybrid silicon detector technology for low energy photons. Inverse Low Gain Avalanche Diode (iLGAD) sensors provide an internal gain, enhancing the signal-to-noise ratio and allowing single photon detection below $1$keV using hybrid detectors. In addition, an optimization of the entrance window of these sensors enhances their quantum efficiency (QE). In this work, the QE and the gain of a batch of different iLGAD diodes with optimized entrance windows were characterized using soft X-rays at the Surface/Interface:Microscopy beamline of the Swiss Light Source synchrotron. Above $250$eV, the QE is larger than $55\%$ for all sensor variations, while the charge collection efficiency is close to $100\%$. The average gain depends on the gain layer design of the iLGADs and increases with photon energy. A fitting procedure is introduced to extract the multiplication factor as a function of the absorption depth of X-ray photons inside the sensors. In particular, the multiplication factors for electron- and hole-triggered avalanches are estimated, corresponding to photon absorption beyond or before the gain layer, respectively., Comment: 16 pages, 8 figures