Search

A likelihood analysis of the observables in deeply virtual exclusive photoproduction off a proton target, $ep \rightarrow e' p' \gamma'$, is presented. Two processes contribute to the reaction: deeply virtual Compton scattering, where the photon is produced at the proton vertex, and the Bether-Heitler process, where the photon is radiated from the electron. We consider the unpolarized process for which the largest amount of data with all the kinematic dependences are available from corresponding datasets with unpolarized beams and unpolarized targets from Jefferson Lab. We provide and use a method which derives a joint likelihood of the Compton form factors, which parametrize the deeply virtual Compton scattering amplitude in QCD, for each observed combination of the kinematic variables defining the reaction. The unpolarized twist-two cross section likelihood fully constrains only three of the Compton form factors (CFFs). The impact of the twist-three corrections to the analysis is also explored. The derived likelihoods are explored using Markov chain Monte Carlo (MCMC) methods. Using our proposed method we derive CFF error bars and covariances. Additionally, we explore methods which may reduce the magnitude of error bars/contours in the future., Comment: 22 pages, 11 figures

Using the equations of motion of QCD and Lorentz invariance relations, we show a new, more direct way of obtaining the decomposition of the proton's quark transverse angular momentum into its spin and orbital components. The new decomposition can be understood as a twist-three relation, with the quark transverse spin described by the parton distribution $g_T$. We find that the orbital component can be expressed in terms of a moment in the quark transverse momentum, $k_T$, of the generalized transverse momentum-dependent distribution $F_{12}$, or alternatively in terms of the twist-three generalized parton distributions $H_{2T}$ and $\widetilde{E}_{2T}$., Comment: 10 pages

In this paper we study the relationships between links in plat position, the dynamics of the braid group, and Heegaard splittings of double branched covers of $S^3$ over a link. These relationships offer new ways to view links in plat position and a new tool kit for analyzing links. In particular, we show that the Hempel distance of the Heegaard splitting of the double branched cover obtained from a plat is a lower bound for the Hempel distance of that plat. Using the Hempel distance of a knot in bridge position and pseudo-Anosov braids we obtain our main result: a construction of infinitely many sequences of prime hyperbolic $n$-bridge knots for $n \geq 3$, infinitely many of which are distinct. We consider known results to show that the knot genus and hyperbolic volume of these knots are bounded below by a linear function. As a further result we show that the plat closure of a 6-braid is generically hyperbolic., Comment: 15 pages, 12 figures, Comments Welcome!

Bridge multisections are combinatorial descriptions of surface links in 4-space using tuples of trivial tangles. They were introduced by Islambouli, Karimi, Lambert-Cole, and Meier to study curves in rational surfaces. In this paper, we prove a uniqueness result for bridge multisections of surfaces in 4-space: we give a complete set of moves relating to any two multiplane diagrams of the same surface. This is done by developing a surgery operation on multiplane diagrams called band surgery. Another application of this surgery move is that any $n$-valent graph with an $n$-edge coloring is the spine of a bridge multisection for an unknotted surface. We also prove that any multisected surface in $S^4$ can be unknotted by finitely many band surgeries., Comment: 23 pages, 25 figures, comments welcome

In this work, we study an inverse reinforcement learning (IRL) problem where the experts are planning under a shared reward function but with different, unknown planning horizons. Without the knowledge of discount factors, the reward function has a larger feasible solution set, which makes it harder for existing IRL approaches to identify a reward function. To overcome this challenge, we develop algorithms that can learn a global multi-agent reward function with agent-specific discount factors that reconstruct the expert policies. We characterize the feasible solution space of the reward function and discount factors for both algorithms and demonstrate the generalizability of the learned reward function across multiple domains., Comment: Accepted at RLC 2024

The stellar masses of galaxies are measured using integrated light via several methods -- however, few of these methods were designed for low-mass ($M_{\star}\lesssim10^{8}\rm{M_{\odot}}$) "dwarf" galaxies, whose properties (e.g., stochastic star formation, low metallicity) pose unique challenges for estimating stellar masses. In this work, we quantify the precision and accuracy at which stellar masses of low-mass galaxies can be recovered using UV/optical/IR photometry. We use mock observations of 469 low-mass galaxies from a variety of models, including both semi-empirical models (GRUMPY, UniverseMachine-SAGA) and cosmological baryonic zoom-in simulations (MARVELous Dwarfs and FIRE-2), to test literature color-$M_\star/L$ relations and multi-wavelength spectral energy distribution (SED) mass estimators. We identify a list of "best practices" for measuring stellar masses of low-mass galaxies from integrated photometry. These include updated prescriptions for stellar mass based on $g-r$ color and WISE 3.4 $\mu$m luminosity, which are less systematically biased than literature calibrations and can recover true stellar masses of low-mass galaxies with $\sim0.1$ dex precision. When using SED fitting to estimate stellar mass, we find that the form of the assumed star formation history can induce significant biases: parametric SFHs can underestimate stellar mass by as much as $\sim0.4$ dex, while non-parametric SFHs recover true stellar masses with insignificant offset ($-0.03\pm0.11$ dex). However, we also caution that non-informative dust attenuation priors may introduce $M_\star$ uncertainties of up to $\sim0.6$ dex., Comment: 28 pages including references, 9 figures; submitted to ApJ

3D reconstruction and relighting of objects made from scattering materials present a significant challenge due to the complex light transport beneath the surface. 3D Gaussian Splatting introduced high-quality novel view synthesis at real-time speeds. While 3D Gaussians efficiently approximate an object's surface, they fail to capture the volumetric properties of subsurface scattering. We propose a framework for optimizing an object's shape together with the radiance transfer field given multi-view OLAT (one light at a time) data. Our method decomposes the scene into an explicit surface represented as 3D Gaussians, with a spatially varying BRDF, and an implicit volumetric representation of the scattering component. A learned incident light field accounts for shadowing. We optimize all parameters jointly via ray-traced differentiable rendering. Our approach enables material editing, relighting and novel view synthesis at interactive rates. We show successful application on synthetic data and introduce a newly acquired multi-view multi-light dataset of objects in a light-stage setup. Compared to previous work we achieve comparable or better results at a fraction of optimization and rendering time while enabling detailed control over material attributes. Project page https://sss.jdihlmann.com/, Comment: Project page: https://sss.jdihlmann.com/

Deeply virtual exclusive scattering processes (DVES) serve as precise probes of nucleon quark and gluon distributions in coordinate space. These distributions are derived from generalized parton distributions (GPDs) via Fourier transform relative to proton momentum transfer. QCD factorization theorems enable DVES to be parameterized by Compton form factors (CFFs), which are convolutions of GPDs with perturbatively calculable kernels. Accurate extraction of CFFs from DVCS, benefiting from interference with the Bethe-Heitler (BH) process and a simpler final state structure, is essential for inferring GPDs. This paper focuses on extracting CFFs from DVCS data using a variational autoencoder inverse mapper (VAIM) and its constrained variant (C-VAIM). VAIM is shown to be consistent with Markov Chain Monte Carlo (MCMC) methods in extracting multiple CFF solutions for given kinematics, while C-VAIM effectively captures correlations among CFFs across different kinematic values, providing more constrained solutions. This study represents a crucial first step towards a comprehensive analysis pipeline towards the extraction of GPDs., Comment: 12 pages, 9 figures

This study investigates the implementation of semi-on-demand (SoD) hybrid-route services using Shared Autonomous Vehicles (SAVs) on existing transit lines. SoD services combine the cost efficiency of fixed-route buses with the flexibility of on-demand services. SAVs first serve all scheduled fixed-route stops, then drop off and pick up passengers in the pre-determined flexible-route portion, and return to the fixed route. This study addresses four key questions: optimal fleet and vehicle sizes for peak-hour fixed-route services with SAVs and during transition (from drivers to autonomous vehicles), optimal off-peak SoD service planning, and suitable use cases. The methodology combines analytical modeling for service planning with agent-based simulation for operational analysis. We examine ten bus routes in Munich, Germany, considering full SAV and transition scenarios with varying proportions of drivers. Our findings demonstrate that the lower operating costs of SAVs improve service quality through increased frequency and smaller vehicles, even in transition scenarios. The reduced headway lowers waiting time and also favors more flexible-route operation in SoD services. The optimal SoD settings range from fully flexible to hybrid routes, where higher occupancy from the terminus favors shorter flexible routes. During the transition phase, limited fleet size and higher headways constrain the benefits of flexible-route operations. The simulation results corroborate the SoD benefits of door-to-door convenience, attracting more passengers without excessive detours and operator costs at moderate flexible-route lengths, and validate the analytical model., Comment: 25 pages, 10 figures

Bringing more transparency to the judicial system for the purposes of increasing accountability often demands extensive effort from auditors who must meticulously sift through numerous disorganized legal case files to detect patterns of bias and errors. For example, the high-profile investigation into the Curtis Flowers case took seven reporters a full year to assemble evidence about the prosecutor's history of selecting racially biased juries. LLMs have the potential to automate and scale these transparency pipelines, especially given their demonstrated capabilities to extract information from unstructured documents. We discuss the opportunities and challenges of using LLMs to provide transparency in two important court processes: jury selection in criminal trials and housing eviction cases., Comment: Accepted at the Seventh AAAI/ACM Conference on AI, Ethics, and Society (AIES 2024)

Rapid urbanization and growing urban populations worldwide present significant challenges for cities, including increased traffic congestion and air pollution. Effective strategies are needed to manage traffic volumes and reduce emissions. In practice, traditional traffic flow simulations are used to test those strategies. However, high computational intensity usually limits their applicability in investigating a magnitude of different scenarios to evaluate best policies. This paper introduces an innovative approach to assess the effects of traffic policies using Graph Neural Networks (GNN). By incorporating complex transport network structures directly into the neural network, this approach could enable rapid testing of various policies without the delays associated with traditional simulations. We provide a proof of concept that GNNs can learn and predict changes in car volume resulting from capacity reduction policies. We train a GNN model based on a training set generated with a MATSim simulation for Paris, France. We analyze the model's performance across different road types and scenarios, finding that the GNN is generally able to learn the effects on edge-based traffic volume induced by policies. The model is especially successful in predicting changes on major streets. Nevertheless, the evaluation also showed that the current model has problems in predicting impacts of spatially small policies and changes in traffic volume in regions where no policy is applied due to spillovers and/or relocation of traffic.

In overview of the recent activity of the newly funded EXCLusives with AI and Machine learning (EXCLAIM) collaboration is presented. The main goal of the collaboration is to develop a framework to implement AI and machine learning techniques in problems emerging from the phenomenology of high energy exclusive scattering processes from nucleons and nuclei, maximizing the information that can be extracted from various sets of experimental data, while implementing theoretical constraints from lattice QCD. A specific perspective embraced by EXCLAIM is to use the methods of theoretical physics to understand the working of ML, beyond its standardized applications to physics analyses which most often rely on industrially provided tools, in an automated way., Comment: 9 pages, 3 figures

Floquet theory and other established semiclassical approaches are widely used methods to predict the state of externally-driven quantum systems, yet, they do not allow to predict the state of the photonic driving field. To overcome this shortcoming, the photon-resolved Floquet theory (PRFT) has been developed recently [Phys. Rev. Research 6, 013116], which deploys concepts from full-counting statistics to predict the statistics of the photon flux between several coherent driving modes. In this paper, we study in detail the scaling properties of the PRFT in the semiclassical regime. We find that there is an ambiguity in the definition of the moment-generating function, such that different versions of the moment-generating function produce the same photonic probability distribution in the semiclassical limit, and generate the same leading-order terms of the moments and cumulants. Using this ambiguity, we establish a simple expression for the Kraus operators, which describe the decoherence dynamics of the driven quantum system appearing as a consequence of the light-matter interaction. The PRFT will pave the way for improved quantum sensing methods, e.g., for spectroscopic quantum sensing protocols, reflectometry in semiconductor nanostructures and other applications, where the detailed knowledge of the photonic probability distribution is necessary., Comment: 17 pages, 4 figures, 3 appendices. Comments are welcome

Contemporary developments in generative AI are rapidly transforming the field of medical AI. These developments have been predominantly driven by the availability of large datasets and high computing power, which have facilitated a significant increase in model capacity. Despite their considerable potential, these models demand substantially high power, leading to high carbon dioxide (CO2) emissions. Given the harm such models are causing to the environment, there has been little focus on the carbon footprints of such models. This study analyzes carbon emissions from 2D and 3D latent diffusion models (LDMs) during training and data generation phases, revealing a surprising finding: the synthesis of large images contributes most significantly to these emissions. We assess different scenarios including model sizes, image dimensions, distributed training, and data generation steps. Our findings reveal substantial carbon emissions from these models, with training 2D and 3D models comparable to driving a car for 10 km and 90 km, respectively. The process of data generation is even more significant, with CO2 emissions equivalent to driving 160 km for 2D models and driving for up to 3345 km for 3D synthesis. Additionally, we found that the location of the experiment can increase carbon emissions by up to 94 times, and even the time of year can influence emissions by up to 50%. These figures are alarming, considering they represent only a single training and data generation phase for each model. Our results emphasize the urgent need for developing environmentally sustainable strategies in generative AI.

A defining property of Hawking radiation is that states with very low entanglement masquerade as highly mixed states; this property is captured by a quantum computational phenomenon known as spoofing entanglement. Motivated by the potential implications for black hole information and the emergence of spacetime connectivity, as well as possible applications of spoofing entanglement, we investigate the geometrization of two types of entanglement spoofers in AdS/CFT: so-called EFI pairs and pseudoentangled state ensembles. We show that (a strengthened version of) EFI pairs with a semiclassical bulk dual have a Python's Lunch; the maximally mixed state over the pseudoentangled state ensemble likewise features a Python's Lunch. Since a Python's Lunch must lie behind an event horizon, we find that black holes are the exclusive gravitational source of entanglement spoofing in the semiclassical limit. Finally, we use an extant construction of holographic pseudorandom states to yield a candidate example of a pseudoentangled state ensemble with a semiclassical bulk dual., Comment: 25+1 pages, 1 figure

The upcoming 6G mobile communication standard will offer a revolutionary new feature: Integrated sensing and communication (ISAC) reuses mobile communication signals to realize multi-static radar for various applications including localization. Consequently, applied ISAC propagation research necessitates to evolve from classical monostatic radar cross section (RCS) measurement of static targets on to bistatic radar reflectivity characterization of dynamic objects. Here, we introduce our "Bistatic Radar" (BIRA) and antenna measurement facility for bistatic spherical positioning with sub-millimeter accuracy on a diameter of up to 7 m and with almost continuous frequency coverage from 0.7 up to 260 GHz. Currently, BIRA is the only bistatic measurement facility capable of unrestricted ISAC research: In addition to vector network analysis, BIRA employs advanced wideband transceiver technology with an instantaneous bandwidth of up to 4 GHz. These transceivers grant BIRA the unique ability to characterize dynamic targets in both Doppler and range, while also significantly accelerating RCS measurements of static objects., Comment: 9 pages, 8 figures

We consider the problem of active learning for global sensitivity analysis of expensive black-box functions. Our aim is to efficiently learn the importance of different input variables, e.g., in vehicle safety experimentation, we study the impact of the thickness of various components on safety objectives. Since function evaluations are expensive, we use active learning to prioritize experimental resources where they yield the most value. We propose novel active learning acquisition functions that directly target key quantities of derivative-based global sensitivity measures (DGSMs) under Gaussian process surrogate models. We showcase the first application of active learning directly to DGSMs, and develop tractable uncertainty reduction and information gain acquisition functions for these measures. Through comprehensive evaluation on synthetic and real-world problems, our study demonstrates how these active learning acquisition strategies substantially enhance the sample efficiency of DGSM estimation, particularly with limited evaluation budgets. Our work paves the way for more efficient and accurate sensitivity analysis in various scientific and engineering applications.

Purpose: Federated training is often hindered by heterogeneous datasets due to divergent data storage options, inconsistent naming schemes, varied annotation procedures, and disparities in label quality. This is particularly evident in the emerging multi-modal learning paradigms, where dataset harmonization including a uniform data representation and filtering options are of paramount importance. Methods: DICOM structured reports enable the standardized linkage of arbitrary information beyond the imaging domain and can be used within Python deep learning pipelines with highdicom. Building on this, we developed an open platform for data integration and interactive filtering capabilities that simplifies the process of assembling multi-modal datasets. Results: In this study, we extend our prior work by showing its applicability to more and divergent data types, as well as streamlining datasets for federated training within an established consortium of eight university hospitals in Germany. We prove its concurrent filtering ability by creating harmonized multi-modal datasets across all locations for predicting the outcome after minimally invasive heart valve replacement. The data includes DICOM data (i.e. computed tomography images, electrocardiography scans) as well as annotations (i.e. calcification segmentations, pointsets and pacemaker dependency), and metadata (i.e. prosthesis and diagnoses). Conclusion: Structured reports bridge the traditional gap between imaging systems and information systems. Utilizing the inherent DICOM reference system arbitrary data types can be queried concurrently to create meaningful cohorts for clinical studies. The graphical interface as well as example structured report templates will be made publicly available.

Federated learning (FL) is a renowned technique for utilizing decentralized data while preserving privacy. However, real-world applications often involve inherent challenges such as partially labeled datasets, where not all clients possess expert annotations of all labels of interest, leaving large portions of unlabeled data unused. In this study, we conduct the largest federated cardiac CT imaging analysis to date, focusing on partially labeled datasets ($n=8,124$) of Transcatheter Aortic Valve Implantation (TAVI) patients over eight hospital clients. Transformer architectures, which are the major building blocks of current foundation models, have shown superior performance when trained on larger cohorts than traditional CNNs. However, when trained on small task-specific labeled sample sizes, it is currently not feasible to exploit their underlying attention mechanism for improved performance. Therefore, we developed a two-stage semi-supervised learning strategy that distills knowledge from several task-specific CNNs (landmark detection and segmentation of calcification) into a single transformer model by utilizing large amounts of unlabeled data typically residing unused in hospitals to mitigate these issues. This method not only improves the predictive accuracy and generalizability of transformer-based architectures but also facilitates the simultaneous learning of all partial labels within a single transformer model across the federation. Additionally, we show that our transformer-based model extracts more meaningful features for further downstream tasks than the UNet-based one by only training the last layer to also solve segmentation of coronary arteries. We make the code and weights of the final model openly available, which can serve as a foundation model for further research in cardiac CT imaging.

Federated learning is one popular paradigm to train a joint model in a distributed, privacy-preserving environment. But partial annotations pose an obstacle meaning that categories of labels are heterogeneous over clients. We propose to learn a joint backbone in a federated manner, while each site receives its own multi-label segmentation head. By using Bayesian techniques we observe that the different segmentation heads although only trained on the individual client's labels also learn information about the other labels not present at the respective site. This information is encoded in their predictive uncertainty. To obtain a final prediction we leverage this uncertainty and perform a weighted averaging of the ensemble of distributed segmentation heads, which allows us to segment "locally unknown" structures. With our method, which we refer to as FUNAvg, we are even on-par with the models trained and tested on the same dataset on average. The code is publicly available at https://github.com/Cardio-AI/FUNAvg., Comment: Accepted at MICCAI24

Quark orbital angular momentum in the proton is evaluated via a Lattice QCD calculation of the second Mellin moment of the twist-3 generalized parton distribution $\widetilde{E}_{2T} $ in the forward limit. The connection between this approach to quark orbital angular momentum and approaches previously utilized in Lattice QCD calculations, via generalized transverse momentum-dependent parton distributions and via Ji's sum rule, is reviewed. This connection can be given in terms of Lorentz invariance and equation of motion relations. The calculation of the second Mellin moment of $\widetilde{E}_{2T} $ proceeds via a finite-momentum proton matrix element of a quark bilocal operator with a straight-line gauge connection and separation in both the longitudinal and transverse directions. The dependence on the former component serves to extract the second Mellin moment, whereas the dependence on the latter component provides a transverse momentum cutoff for the matrix element. Furthermore, a derivative of the matrix element with respect to momentum transfer in the forward limit is required, which is obtained using a direct derivative method. The calculation utilizes a clover fermion ensemble at pion mass 317 MeV. The resulting quark orbital angular momentum is consistent with previous evaluations through alternative approaches, albeit with greater statistical uncertainty using a comparable number of samples., Comment: 10 pages, 3 figures, to appear in the proceedings of the 25th International Spin Physics Symposium (SPIN 2023), PoS SPIN2023 (2024) 075

In recent times, large language models (LLMs) have made significant strides in generating computer code, blurring the lines between code created by humans and code produced by artificial intelligence (AI). As these technologies evolve rapidly, it is crucial to explore how they influence code generation, especially given the risk of misuse in areas like higher education. This paper explores this issue by using advanced classification techniques to differentiate between code written by humans and that generated by ChatGPT, a type of LLM. We employ a new approach that combines powerful embedding features (black-box) with supervised learning algorithms - including Deep Neural Networks, Random Forests, and Extreme Gradient Boosting - to achieve this differentiation with an impressive accuracy of 98%. For the successful combinations, we also examine their model calibration, showing that some of the models are extremely well calibrated. Additionally, we present white-box features and an interpretable Bayes classifier to elucidate critical differences between the code sources, enhancing the explainability and transparency of our approach. Both approaches work well but provide at most 85-88% accuracy. We also show that untrained humans solve the same task not better than random guessing. This study is crucial in understanding and mitigating the potential risks associated with using AI in code generation, particularly in the context of higher education, software development, and competitive programming., Comment: Accepted for publication in MDPI AI Journal

We propose 3D Congealing, a novel problem of 3D-aware alignment for 2D images capturing semantically similar objects. Given a collection of unlabeled Internet images, our goal is to associate the shared semantic parts from the inputs and aggregate the knowledge from 2D images to a shared 3D canonical space. We introduce a general framework that tackles the task without assuming shape templates, poses, or any camera parameters. At its core is a canonical 3D representation that encapsulates geometric and semantic information. The framework optimizes for the canonical representation together with the pose for each input image, and a per-image coordinate map that warps 2D pixel coordinates to the 3D canonical frame to account for the shape matching. The optimization procedure fuses prior knowledge from a pre-trained image generative model and semantic information from input images. The former provides strong knowledge guidance for this under-constraint task, while the latter provides the necessary information to mitigate the training data bias from the pre-trained model. Our framework can be used for various tasks such as correspondence matching, pose estimation, and image editing, achieving strong results on real-world image datasets under challenging illumination conditions and on in-the-wild online image collections., Comment: Project page: https://ai.stanford.edu/~yzzhang/projects/3d-congealing/

This paper aims to improve the average response time for naval accidents in the North and Baltic Sea. To do this we optimize the strategic distribution of the vessel fleet used by the Deutsche Gesellschaft zur Rettung Schiffbr\"uchiger (German Maritime Search and Rescue Service) (DGzRS) across several home stations. Based on these locations, in case of an incoming distress call the vessel with the lowest response time is dispatched. A particularity of the region considered is the fact that due to low tide, at predictable times some vessels and stations are not operational. In our work, we build a corresponding mathematical model for the allocation of rescue crafts to multiple stations. Thereafter, we show that the problem is NP-hard. Next, we provide an Integer Programming (IP) formulation. Finally, we propose several methods of simplifying the model and do a case study to compare their effectiveness. For this, we generate test instances based on real-world data., Comment: Code URL: https://github.com/veni-vidi-code/RCA

The flow convergence method includes calculation of the proximal isovelocity surface area (PISA) and is widely used to classify mitral regurgitation (MR) with echocardiography. It constitutes a primary decision factor for determination of treatment and should therefore be a robust quantification method. However, it is known for its tendency to underestimate MR and its dependence on user expertise. The present work systematically compares different pulsatile flow profiles arising from different regurgitation orifices using transesophageal echocardiographic (TEE) probe and particle image velocimetry (PIV) as a reference in an in-vitro environment. It is found that the inter-observer variability using echocardiography is small compared to the systematic underestimation of the regurgitation volume for large orifice areas (up to 52%) where a violation of the flow convergence method assumptions occurs. From a flow perspective, a starting vortex was found as a dominant flow pattern in the regurgant jet for all orifice shapes and sizes. A series of simplified computational fluid dynamics (CFD) simulations indicate that selecting a suboptimal aliasing velocity during echocardiography measurements might be a primary source of potential underestimation in MR characterization via the PISA-based method, reaching up to 40%. In this study, it has been noted in clinical observations that physicians often select an aliasing velocity higher than necessary for optimal estimation in diagnostic procedures.

Visual navigation requires a whole range of capabilities. A crucial one of these is the ability of an agent to determine its own location and heading in an environment. Prior works commonly assume this information as given, or use methods which lack a suitable inductive bias and accumulate error over time. In this work, we show how the method of slow feature analysis (SFA), inspired by neuroscience research, overcomes both limitations by generating interpretable representations of visual data that encode location and heading of an agent. We employ SFA in a modern reinforcement learning context, analyse and compare representations and illustrate where hierarchical SFA can outperform other feature extractors on navigation tasks., Comment: Accepted at XAI4DRL workshop at AAAI 2024

We present deep Hubble Space Telescope (HST) photometry of ten targets from Treasury Program GO-14734, including six confirmed ultra-faint dwarf galaxies (UFDs), three UFD candidates, and one likely globular cluster. Six of these targets are satellites of, or have interacted with, the Large Magellanic Cloud (LMC). We determine their structural parameters using a maximum-likelihood technique. Using our newly derived half-light radius ($r_h$) and $V$-band magnitude ($M_V$) values in addition to literature values for other UFDs, we find that UFDs associated with the LMC do not show any systematic differences from Milky Way UFDs in the magnitude-size plane. Additionally, we convert simulated UFD properties from the literature into the $M_V-r_h$ observational space to examine the abilities of current dark matter (DM) and baryonic simulations to reproduce observed UFDs. Some of these simulations adopt alternative DM models, thus allowing us to also explore whether the $M_V-r_h$ plane could be used to constrain the nature of DM. We find no differences in the magnitude-size plane between UFDs simulated with cold, warm, and self-interacting dark matter, but note that the sample of UFDs simulated with alternative DM models is quite limited at present. As more deep, wide-field survey data become available, we will have further opportunities to discover and characterize these ultra-faint stellar systems and the greater low surface-brightness universe., Comment: 42 pages, 13 figures, 13 tables, 2 appendices; Submitted to AAS Journals

The upcoming 3GPP global mobile communication standard 6G strives to push the technological limits of radio frequency (RF) communication even further than its predecessors: Sum data rates beyond 100 Gbit/s, RF bandwidths above 1 GHz per link, and sub-millisecond latency necessitate very high performance development tools. We propose a new SDR firmware and software architecture designed explicitly to meet these challenging requirements. It relies on Ethernet and commercial off-the-shelf network and server components to maximize flexibility and to reduce costs. We analyze state-of-the-art solutions (USRP X440 and other RFSoC-based systems), derive architectural design goals, explain resulting design decision in detail, and exemplify our architecture's implementation on the XCZU48DR RFSoC. Finally, we validate its performance via measurements and outline how the architecture surpasses the state-of-the-art with respect to sustained RF recording, while maintaining high Ethernet bandwidth efficiency. Building a 6G integrated sensing and communication (ISAC) example, we demonstrate its real-time and rapid application development capabilities.

We formulate and take two large strides towards proving a quantum version of the weak cosmic censorship conjecture. We first prove "Cryptographic Censorship": a theorem showing that when the time evolution operator of a holographic CFT is approximately pseudorandom (or Haar random) on some code subspace, then there must be an event horizon in the corresponding bulk dual. This result provides a general condition that guarantees (in finite time) event horizon formation, with minimal assumptions about the global spacetime structure. Our theorem relies on an extension of a recent quantum learning no-go theorem and is proved using new techniques of pseudorandom measure concentration. To apply this result to cosmic censorship, we separate singularities into classical, semi-Planckian, and Planckian types. We illustrate that classical and semi-Planckian singularities are compatible with approximately pseudorandom CFT time evolution; thus, if such singularities are indeed approximately pseudorandom, by Cryptographic Censorship, they cannot exist in the absence of event horizons. This result provides a sufficient condition guaranteeing that seminal holographic results on quantum chaos and thermalization, whose general applicability relies on typicality of horizons, will not be invalidated by the formation of naked singularities in AdS/CFT., Comment: 50 pages + appendices, 15 figures; v2: minor edits, additional explanations added

AI models present a wide range of applications in the field of medicine. However, achieving optimal performance requires access to extensive healthcare data, which is often not readily available. Furthermore, the imperative to preserve patient privacy restricts patient data sharing with third parties and even within institutes. Recently, generative AI models have been gaining traction for facilitating open-data sharing by proposing synthetic data as surrogates of real patient data. Despite the promise, these models are susceptible to patient data memorization, where models generate patient data copies instead of novel synthetic samples. Considering the importance of the problem, it has received little attention in the medical imaging community. To this end, we assess memorization in unconditional latent diffusion models. We train 2D and 3D latent diffusion models on CT, MR, and X-ray datasets for synthetic data generation. Afterwards, we detect the amount of training data memorized utilizing our self-supervised approach and further investigate various factors that can influence memorization. Our findings show a surprisingly high degree of patient data memorization across all datasets, with approximately 40.9% of patient data being memorized and 78.5% of synthetic samples identified as patient data copies on average in our experiments. Further analyses reveal that using augmentation strategies during training can reduce memorization while over-training the models can enhance it. Although increasing the dataset size does not reduce memorization and might even enhance it, it does lower the probability of a synthetic sample being a patient data copy. Collectively, our results emphasize the importance of carefully training generative models on private medical imaging datasets, and examining the synthetic data to ensure patient privacy before sharing it for medical research and applications.

We present SHINOBI, an end-to-end framework for the reconstruction of shape, material, and illumination from object images captured with varying lighting, pose, and background. Inverse rendering of an object based on unconstrained image collections is a long-standing challenge in computer vision and graphics and requires a joint optimization over shape, radiance, and pose. We show that an implicit shape representation based on a multi-resolution hash encoding enables faster and robust shape reconstruction with joint camera alignment optimization that outperforms prior work. Further, to enable the editing of illumination and object reflectance (i.e. material) we jointly optimize BRDF and illumination together with the object's shape. Our method is class-agnostic and works on in-the-wild image collections of objects to produce relightable 3D assets for several use cases such as AR/VR, movies, games, etc. Project page: https://shinobi.aengelhardt.com Video: https://www.youtube.com/watch?v=iFENQ6AcYd8&feature=youtu.be, Comment: Accepted by IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR 2024). Updated supplementary material and acknowledgements

We use high-resolution cosmological simulations to compare the effect of bursty star formation histories on dwarf galaxy structure for two different subgrid supernovae (SNe) feedback models in dwarf galaxies with stellar masses from $5000 <$ M$_*$/M$_\odot$ $< 10^{9}$. Our simulations are run using two distinct supernova feedback models: superbubble and blastwave. We show that both models are capable of producing galaxies that are cored and reproduce observed scaling relations for metallicity, luminosity, mass, and size. We show that continuous bursty star formation and the resulting stellar feedback are able to sustain dark matter cores in the higher dwarf galaxy mass regime, while the majority of ultra-faint and classical dwarfs retain cuspy central dark matter density profiles. We find that both subgrid SN models are able to create bursty star formation histories. We find that effective core formation peaks at M$_*$/M$_\odot$ $\simeq 5 \times 10^{-3}$ for both feedback models. Galaxies simulated with superbubble feedback peak at lower mean burstiness values relative to blastwave feedback, indicating that core formation in the superbubble sample may be less motivated by the burstiness of star formation., Comment: 18 pages, 7 figures

Contrastive dimension reduction methods have been developed for case-control study data to identify variation that is enriched in the foreground (case) data X relative to the background (control) data Y. Here, we develop contrastive regression for the setting when there is a response variable r associated with each foreground observation. This situation occurs frequently when, for example, the unaffected controls do not have a disease grade or intervention dosage but the affected cases have a disease grade or intervention dosage, as in autism severity, solid tumors stages, polyp sizes, or warfarin dosages. Our contrastive regression model captures shared low-dimensional variation between the predictors in the cases and control groups, and then explains the case-specific response variables through the variance that remains in the predictors after shared variation is removed. We show that, in one single-nucleus RNA sequencing dataset on autism severity in postmortem brain samples from donors with and without autism and in another single-cell RNA sequencing dataset on cellular differentiation in chronic rhinosinusitis with and without nasal polyps, our contrastive linear regression performs feature ranking and identifies biologically-informative predictors associated with response that cannot be identified using other approaches

Advances in image diffusion models have recently led to notable improvements in the generation of high-quality images. In combination with Neural Radiance Fields (NeRFs), they enabled new opportunities in 3D generation. However, most generative 3D approaches are object-centric and applying them to editing existing photorealistic scenes is not trivial. We propose SIGNeRF, a novel approach for fast and controllable NeRF scene editing and scene-integrated object generation. A new generative update strategy ensures 3D consistency across the edited images, without requiring iterative optimization. We find that depth-conditioned diffusion models inherently possess the capability to generate 3D consistent views by requesting a grid of images instead of single views. Based on these insights, we introduce a multi-view reference sheet of modified images. Our method updates an image collection consistently based on the reference sheet and refines the original NeRF with the newly generated image set in one go. By exploiting the depth conditioning mechanism of the image diffusion model, we gain fine control over the spatial location of the edit and enforce shape guidance by a selected region or an external mesh., Comment: Project Page: https://signerf.jdihlmann.com

Broadband infrastructure owners do not always know how their customers are connected in the local networks, which are structured as rooted trees. A recent study is able to infer the topology of a local network using discrete time series data from the leaves of the tree (customers). In this study we propose a contrastive approach for learning a binary event encoder from continuous time series data. As a preliminary result, we show that our approach has some potential in learning a valuable encoder., Comment: Extended abstract presented as a poster at the Northern Lights Deep Learning Conference 2024 in Troms{\o}, Norway

Patient-to-room assignment (PRA) is a scheduling problem in decision support for hospitals. It consists of assigning patients to rooms according to certain objectives, e.g., avoiding transfers and respecting single-room requests. This work presents combinatorial insights about the feasibility of PRA and about the assignment of patients to single rooms. We further compare different IP-formulations for PRA as well as the influence of different objectivs on the runtime. Based on these results, we develop a fast IP-based solution approach which obtains high quality solution. The applicability is verified through a computational study with instances derived from real-world data. Results indicate that large, real world instances can be solved to a high degree of optimality within (fractions of) seconds.

The second Mellin moments $\langle x\rangle$ of the nucleon's unpolarized, polarized, and transversity parton distribution functions (PDFs) are computed. Two lattice QCD ensembles at the physical pion mass are used: these were generated using a tree-level Symanzik-improved gauge action and 2+1 flavour tree-level improved Wilson Clover fermions coupling via 2-level HEX-smearing. The moments are extracted from forward matrix elements of local leading twist operators. We determine renomalization factors in RI-(S)MOM and match to $\overline{\mathrm{MS}}$ at scale $2\,\mathrm{GeV}$. Our findings show that operators that exhibit vanishing kinematics at zero momentum can have significantly reduced excited-state contamination. The resulting polarized moment is used to quantify the longitudinal contribution to the quark spin-orbit correlation. All our results agree within two sigma with previous lattice results.