Search

Evaluating spatial patterns in data is an integral task across various domains, including geostatistics, astronomy, and spatial tissue biology. The analysis of transcriptomics data in particular relies on methods for detecting spatially-dependent features that exhibit significant spatial patterns for both explanatory analysis and feature selection. However, given the complex and high-dimensional nature of these data, there is a need for robust, stable, and reliable descriptors of spatial dependence. We leverage the stability and multiscale properties of persistent homology to address this task. To this end, we introduce a novel framework using functional topological summaries, such as Betti curves and persistence landscapes, for identifying and describing non-random patterns in spatial data. In particular, we propose a non-parametric one-sample permutation test for spatial dependence and investigate its utility across both simulated and real spatial omics data. Our vectorised approach outperforms baseline methods at accurately detecting spatial dependence. Further, we find that our method is more robust to outliers than alternative tests using Moran's I.

Real-world processes often exhibit temporal separation between actions and reactions - a characteristic frequently ignored in many modelling frameworks. Adding temporal aspects, like time delays, introduces a higher complexity of problems and leads to models that are challenging to analyse and computationally expensive to solve. In this work, we propose an intermediate solution to resolve the issue in the framework of evolutionary game theory. Our compartment-based model includes time delays while remaining relatively simple and straightforward to analyse. We show that this model yields qualitatively comparable results with models incorporating explicit delays. Particularly, we focus on the case of delays between parents' interaction and an offspring joining the population, with the magnitude of the delay depending on the parents' strategy. We analyse Stag-Hunt, Snowdrift, and the Prisoner's Dilemma game and show that strategy-dependent delays are detrimental to affected strategies. Additionally, we present how including delays may change the effective games played in the population, subsequently emphasising the importance of considering the studied systems' temporal aspects to model them accurately.

This paper introduces the sigma flow model for the prediction of structured labelings of data observed on Riemannian manifolds, including Euclidean image domains as special case. The approach combines the Laplace-Beltrami framework for image denoising and enhancement, introduced by Sochen, Kimmel and Malladi about 25 years ago, and the assignment flow approach introduced and studied by the authors. The sigma flow arises as Riemannian gradient flow of generalized harmonic energies and thus is governed by a nonlinear geometric PDE which determines a harmonic map from a closed Riemannian domain manifold to a statistical manifold, equipped with the Fisher-Rao metric from information geometry. A specific ingredient of the sigma flow is the mutual dependency of the Riemannian metric of the domain manifold on the evolving state. This makes the approach amenable to machine learning in a specific way, by realizing this dependency through a mapping with compact time-variant parametrization that can be learned from data. Proof of concept experiments demonstrate the expressivity of the sigma flow model and prediction performance. Structural similarities to transformer network architectures and networks generated by the geometric integration of sigma flows are pointed out, which highlights the connection to deep learning and, conversely, may stimulate the use of geometric design principles for structured prediction in other areas of scientific machine learning., Comment: 51 pages

Advancements in artificial intelligence have driven the development of numerous pathology foundation models capable of extracting clinically relevant information. However, there is currently limited literature independently evaluating these foundation models on truly external cohorts and clinically-relevant tasks to uncover adjustments for future improvements. In this study, we benchmarked ten histopathology foundation models on 13 patient cohorts with 6,791 patients and 9,493 slides from lung, colorectal, gastric, and breast cancers. The models were evaluated on weakly-supervised tasks related to biomarkers, morphological properties, and prognostic outcomes. We show that a vision-language foundation model, CONCH, yielded the highest performance in 42% of tasks when compared to vision-only foundation models. The experiments reveal that foundation models trained on distinct cohorts learn complementary features to predict the same label, and can be fused to outperform the current state of the art. Creating an ensemble of complementary foundation models outperformed CONCH in 66% of tasks. Moreover, our findings suggest that data diversity outweighs data volume for foundation models. Our work highlights actionable adjustments to improve pathology foundation models.

Identifying (a) systemic barriers to quality healthcare access and (b) key indicators of care efficacy in the United States remains a significant challenge. To improve our understanding of regional disparities in care delivery, we introduce a novel application of curvature, a geometrical-topological property of networks, to Physician Referral Networks. Our initial findings reveal that Forman-Ricci and Ollivier-Ricci curvature measures, which are known for their expressive power in characterizing network structure, offer promising indicators for detecting variations in healthcare efficacy while capturing a range of significant regional demographic features. We also present APPARENT, an open-source tool that leverages Ricci curvature and other network features to examine correlations between regional Physician Referral Networks structure, local census data, healthcare effectiveness, and patient outcomes.

In the rapidly evolving field of cybersecurity, the integration of flow-level and packet-level information for real-time intrusion detection remains a largely untapped area of research. This paper introduces "XG-NID," a novel framework that, to the best of our knowledge, is the first to fuse flow-level and packet-level data within a heterogeneous graph structure, offering a comprehensive analysis of network traffic. Leveraging a heterogeneous graph neural network (GNN) with graph-level classification, XG-NID uniquely enables real-time inference while effectively capturing the intricate relationships between flow and packet payload data. Unlike traditional GNN-based methodologies that predominantly analyze historical data, XG-NID is designed to accommodate the heterogeneous nature of network traffic, providing a robust and real-time defense mechanism. Our framework extends beyond mere classification; it integrates Large Language Models (LLMs) to generate detailed, human-readable explanations and suggest potential remedial actions, ensuring that the insights produced are both actionable and comprehensible. Additionally, we introduce a new set of flow features based on temporal information, further enhancing the contextual and explainable inferences provided by our model. To facilitate practical application and accessibility, we developed "GNN4ID," an open-source tool that enables the extraction and transformation of raw network traffic into the proposed heterogeneous graph structure, seamlessly integrating flow and packet-level data. Our comprehensive quantitative comparative analysis demonstrates that XG-NID achieves an F1 score of 97\% in multi-class classification, outperforming existing baseline and state-of-the-art methods. This sets a new standard in Network Intrusion Detection Systems by combining innovative data fusion with enhanced interpretability and real-time capabilities., Comment: 19 pages, 6 figures

Speech Emotion Recognition (SER) needs high computational resources to overcome the challenge of substantial annotator disagreement. Today SER is shifting towards dimensional annotations of arousal, dominance, and valence (A/D/V). Universal metrics as the L2 distance prove unsuitable for evaluating A/D/V accuracy due to non converging consensus of annotator opinions. However, Concordance Correlation Coefficient (CCC) arose as an alternative metric for A/D/V where a model's output is evaluated to match a whole dataset's CCC rather than L2 distances of individual audios. Recent studies have shown that wav2vec2 / wavLM architectures outputing a float value for each A/D/V dimension achieve today's State-of-the-art (Sota) CCC on A/D/V. The Wav2Vec2.0 / WavLM family has a high computational footprint, but training small models using human annotations has been unsuccessful. In this paper we use a large Transformer Sota A/D/V model as Teacher/Annotator to train 5 student models: 4 MobileNets and our proposed Wav2Small, using only the Teacher's A/D/V outputs instead of human annotations. The Teacher model we propose also sets a new Sota on the MSP Podcast dataset of valence CCC=0.676. We choose MobileNetV4 / MobileNet-V3 as students, as MobileNet has been designed for fast execution times. We also propose Wav2Small - an architecture designed for minimal parameters and RAM consumption. Wav2Small with an .onnx (quantised) of only 120KB is a potential solution for A/D/V on hardware with low resources, having only 72K parameters vs 3.12M parameters for MobileNet-V4-Small., Comment: Nomenclature

This paper deals with 3D reconstruction of seabirds which recently came into focus of environmental scientists as valuable bio-indicators for environmental change. Such 3D information is beneficial for analyzing the bird's behavior and physiological shape, for example by tracking motion, shape, and appearance changes. From a computer vision perspective birds are especially challenging due to their rapid and oftentimes non-rigid motions. We propose an approach to reconstruct the 3D pose and shape from monocular videos of a specific breed of seabird - the common murre. Our approach comprises a full pipeline of detection, tracking, segmentation, and temporally consistent 3D reconstruction. Additionally, we propose a temporal loss that extends current single-image 3D bird pose estimators to the temporal domain. Moreover, we provide a real-world dataset of 10000 frames of video observations on average capture nine birds simultaneously, comprising a large variety of motions and interactions, including a smaller test set with bird-specific keypoint labels. Using our temporal optimization, we achieve state-of-the-art performance for the challenging sequences in our dataset.

All-optical helicity-independent switching (AO-HIS) is of interest for ultrafast and energy efficient magnetic switching in future magnetic data storage approaches. Yet, to achieve high bit density magnetic recording it is necessary to reduce the size of the magnetic bits addressed by laser pulses at well-controlled positions. Metallic nanostructures that support localized surface plasmons enable spatial electromagnetic confinement well below the diffraction limit and rare-earth transition metal alloys such as GdTbCo have demonstrated nanometre-sized stable domains. Here, we deposit plasmonic gold nanostructures on a GdTbCo film and probe the magnetic state using magnetic force microscopy. We observe localized AO-HIS down to a critical dimension of 240 nm after excitation of the gold nanostructures by a single 370 fs long laser pulse with a centre wavelength of 1030 nm. We demonstrate that the strong localization of optical fields through plasmonic nanostructures enables reproducible localized nanoscale AO-HIS at sub-wavelength length scales. We study the influence of the localized electromagnetic field enhancement by the plasmonic nanostructures on the required fluence to switch the magnetization., Comment: 10 pages, 4 figures

Particle-particle and particle-gas processes significantly impact planetary precursors such as dust aggregates and planetesimals. We investigate gas permeability ($\kappa$) in 12 granular samples, mimicking planetesimal dust regoliths. Using parabolic flights, this study assesses how gravitational compression -- and lack thereof -- influences gas permeation, impacting the equilibrium state of low-gravity objects. Transitioning between micro- and hyper-gravity induces granular sedimentation dynamics, revealing collective dust-grain aerodynamics. Our experiments measure $\kappa$ across Knudsen number (Kn) ranges, reflecting transitional flow. Using mass and momentum conservation, we derive $\kappa$ and calculate pressure gradients within the granular matrix. Key findings: 1. As confinement pressure increases with gravitational load and mass flow, $\kappa$ and average pore space decrease. This implies that a planetesimal's unique dust-compaction history limits sub-surface volatile outflows. 2. The derived pressure gradient enables tensile strength determination for asteroid regolith simulants with cohesion. This offers a unique approach to studying dust-layer properties when suspended in confinement pressures comparable to the equilibrium state on planetesimals surfaces, which will be valuable for modelling their collisional evolution. 3. We observe a dynamical flow symmetry breaking when granular material moves against the pressure gradient. This occurs even at low Reynolds numbers, suggesting that Stokes numbers for drifting dust aggregates near the Stokes-Epstein transition require a drag force modification based on permeability.

Persistent homology is one of the most popular methods in Topological Data Analysis. An initial step in any analysis with persistent homology involves constructing a nested sequence of simplicial complexes, called a filtration, from a point cloud. There is an abundance of different complexes to choose from, with Rips, Alpha, and witness complexes being popular choices. In this manuscript, we build a different type of a geometrically-informed simplicial complex, called an ellipsoid complex. This complex is based on the idea that ellipsoids aligned with tangent directions better approximate the data compared to conventional (Euclidean) balls centered at sample points that are used in the construction of Rips and Alpha complexes, for instance. We use Principal Component Analysis to estimate tangent spaces directly from samples and present algorithms as well as an implementation for computing ellipsoid barcodes, i.e., topological descriptors based on ellipsoid complexes. Furthermore, we conduct extensive experiments and compare ellipsoid barcodes with standard Rips barcodes. Our findings indicate that ellipsoid complexes are particularly effective for estimating homology of manifolds and spaces with bottlenecks from samples. In particular, the persistence intervals corresponding to a ground-truth topological feature are longer compared to the intervals obtained when using the Rips complex of the data. Furthermore, ellipsoid barcodes lead to better classification results in sparsely-sampled point clouds. Finally, we demonstrate that ellipsoid barcodes outperform Rips barcodes in classification tasks.

We investigate the emergence of helical edge modes in a Heisenberg antiferromagnet on a triangular lattice, driven by a topological mechanism similar to that proposed by Dong et al. [Phys. Rev. Lett. 130, 206701 (2023)] for chiral spin waves in ferromagnets. The spin-frame field theory of a three-sublattice antiferromagnet allows for a topological term in the energy that modifies the boundary conditions for certain polarizations of spin waves and gives rise to edge modes. These edge modes are helical: modes with left and right circular polarizations propagate in opposite directions along the boundary in a way reminiscent of the electron edge modes in two-dimensional topological insulators. The field-theoretic arguments are verified in a realistic lattice model of a Heisenberg antiferromagnet with superexchange interactions that exhibits helical edge modes. The strength of the topological term is proportional to the disparity between two inequivalent superexchange paths. These findings suggest potential avenues for realizing magnonic edge states in frustrated antiferromagnets without requiring Dzyaloshinskii-Moriya interactions or nontrivial magnon band topology., Comment: 10 pages, 6 figures

This work aims to analyze some of the polluters proposed in the self-enrichment scenarios put forward to explain the multiple populations in globular clusters (GCs), extending previous studies. Three scenarios with different polluter stars were tested: asymptotic giant branch stars, high-mass interacting binaries, and fast rotating massive stars. With abundance data available from the APOGEE survey and $\Delta Y$ estimates from precise HST photometry, twenty-six clusters were studied. We also included the study of the abundances of N, C, Mg and Al, extending previous studies that focused mainly on the abundances of He, O and Na. In addition, we constructed an empirical model to test whether one could explain the chemical signatures of the 'enriched' population of GC stars with a fixed source and dilution process based on empirical data. In agreement with work by other authors, we found that the proposed polluters can generally predict the qualitative abundance patterns in GC stars and in some cases quantitatively predict some elements, but in most cases when we compare the model yields with the observations, we find that they can not explain the entire set of observed abundance patterns. The empirical model succeeds in reproducing the abundances of Al for a given $\Delta Y$ (and vice versa), showing that there is a direct relationship between Al and He, with one increasing proportionally to the other. However, the empirical model fails to reproduce the observed abundances of Na and N, in agreement with the results of previous works. The observed decoupling between the maximum abundances of CNO-cycle elements like N and Na with those of Al and He provides new information and constraints for future models and could take us a step closer to understanding the origin of the peculiar abundance variations of globular cluster stars.

The formation of massive objects via gravitational collapse is relevant both for explaining the origin of the first supermassive black holes and in the context of massive star formation. Here, we analyze simulations of the formation of massive objects pursued by different groups and in various environments, concerning the formation of supermassive black holes, primordial stars, as well as present-day massive stars. We focus particularly on the regime of small virial parameters, i.e., low ratios of the initial kinetic to gravitational energy, low to moderate Mach numbers, and the phase before feedback is very efficient. We compare the outcomes of collapse under different conditions using dimensionless parameters, particularly the star formation efficiency \epsilon_*, the fraction f_* of mass in the most massive object relative to the total stellar mass, and the fraction f_{\rm tot} of mass of the most massive object as a function of the total mass. We find that in all simulations analyzed here, f_{\rm tot} increases as a function of \epsilon_*, although the steepness of the increase depends on the environment. The relation between f_* and \epsilon_* is found to be more complex and also strongly depends on the number of protostars present at the beginning of the simulations. We show that a collision parameter, estimated as the ratio of the system size divided by the typical collision length, allows us to approximately characterize whether collisions will be important. We analyze the statistical correlation between the dimensionless quantities using the Spearman coefficient and confirm via a machine learning analysis that good predictions of f_* can be obtained from \epsilon_* together with a rough estimate of the collision parameter. This suggests that a good estimate of the mass of the most massive object can be obtained once the maximum efficiency for a given environment is known., Comment: 13 pages, 8 figures, 2 tables. Submitted to A&A

This work introduces Dareplane, a modular and broad technology agnostic open source software platform for brain-computer interface research with an application focus on adaptive deep brain stimulation (aDBS). While the search for suitable biomarkers to inform aDBS has provided rich results over the last two decades, development of control strategies is not progressing at the same pace. One difficulty for investigating control approaches resides with the complex setups required for aDBS experiments. The Dareplane platform supports aDBS setups, and more generally brain computer interfaces, by providing a modular, technology-agnostic, and easy-to-implement software platform to make experimental setups more resilient and replicable. The key features of the platform are presented and the composition of modules into a full experimental setup is discussed in the context of a Python-based orchestration module. The performance of a typical experimental setup on Dareplane for aDBS is evaluated in three benchtop experiments, covering (a) an easy-to-replicate setup using an Arduino microcontroller, (b) a setup with hardware of an implantable pulse generator, and (c) a setup using an established and CE certified external neurostimulator. Benchmark results are presented for individual processing steps and full closed-loop processing. The results show that the microcontroller setup in (a) provides timing comparable to the realistic setups in (b) and (c). The Dareplane platform was successfully used in a total of 19 open-loop DBS sessions with externalized DBS and electrocorticography (ECoG) leads. In addition, the full technical feasibility of the platform in the aDBS context is demonstrated in a first closed-loop session with externalized leads on a patient with Parkinson's disease receiving DBS treatment.

In this work, we introduce Gemma 2, a new addition to the Gemma family of lightweight, state-of-the-art open models, ranging in scale from 2 billion to 27 billion parameters. In this new version, we apply several known technical modifications to the Transformer architecture, such as interleaving local-global attentions (Beltagy et al., 2020a) and group-query attention (Ainslie et al., 2023). We also train the 2B and 9B models with knowledge distillation (Hinton et al., 2015) instead of next token prediction. The resulting models deliver the best performance for their size, and even offer competitive alternatives to models that are 2-3 times bigger. We release all our models to the community.

Controlling light propagation in photonic systems fosters fundamental research and practical application. Particularly, photonic lattices allow engineering band dispersions and tailor transport features through their geometry. However, complete controllability requires external manipulation of the propagating light. Here, we present a resonant excitation scheme to observe quasi-1D and uni-directional propagation of light through the bulk of two-dimensional lattices. To this end, we use the highly anisotropic light propagation exhibited at the energy of saddle points in photonic bands. When multiple drives with judicious amplitudes and phases are tuned to such energy, interference effects between these drives and photonic modes result in controllable directional propagation through the bulk. Similarly, one can formed localized states with controllable localization degrees. We illustrate these effects with driven-dissipative photonic lattices. Our work highlights the importance of external drives for dynamically controlling directional light transport in lattices, a relevant feature for all-optical routing and processing in photonics., Comment: 7 pages, 5 figures

We derive a central limit theorem for a spatial $\Lambda$-Fleming-Viot model with fluctuating population size. At each reproduction, a proportion of the population dies and is replaced by a not necessarily equal mass of new individuals. The mass depends on the local population size and a function thereof. Additionally, as new individuals have a single parental type, with growing population size, events become more frequent and of smaller impact, modelling the successful reproduction of a higher number of individuals. From the central limit theorem we derive a Wright-Mal\'ecot formula quantifying the asymptotic probability of identity by descent and thus isolation by distance. The formula reflects that ancestral lineages are attracted by centres of population mass and coalesce with a rate inversely proportional to the population size. Notably, we obtain this information despite the varying population size rendering the dual process intractable., Comment: 54 pages, 2 figures

The mechanics of animal cells is strongly determined by stress fibers, which are contractile filament bundles that form dynamically in response to extracellular cues. Stress fibers allow the cell to adapt its mechanics to environmental conditions and to protect it from structural damage. While the physical description of single stress fibers is well-developed, much less is known about their spatial distribution on the level of whole cells. Here, we combine a finite element method for one-dimensional fibers embedded in an elastic bulk medium with dynamical rules for stress fiber formation based on genetic algorithms. We postulate that their main goal is to achieve minimal mechanical stress in the bulk material with as few fibers as possible. The fiber positions and configurations resulting from this optimization task alone are in good agreement with those found in experiments where cells in 3D-scaffolds were mechanically strained at one attachment point. For optimized configurations, we find that stress fibers typically run through the cell in a diagonal fashion, similar to reinforcement strategies used for composite material., Comment: 33 pages, 10 figures, 64 references

We consider the problem of detecting gradual changes in the sequence of mean functions from a not necessarily stationary functional time series. Our approach is based on the maximum deviation (calculated over a given time interval) between a benchmark function and the mean functions at different time points. We speak of a gradual change of size $\Delta $, if this quantity exceeds a given threshold $\Delta>0$. For example, the benchmark function could represent an average of yearly temperature curves from the pre-industrial time, and we are interested in the question if the yearly temperature curves afterwards deviate from the pre-industrial average by more than $\Delta =1.5$ degrees Celsius, where the deviations are measured with respect to the sup-norm. Using Gaussian approximations for high-dimensional data we develop a test for hypotheses of this type and estimators for the time where a deviation of size larger than $\Delta$ appears for the first time. We prove the validity of our approach and illustrate the new methods by a simulation study and a data example, where we analyze yearly temperature curves at different stations in Australia.

Double beta plus decay is a rare nuclear disintegration process. Difficulties in its measurement arise from suppressed decay probabilities, experimentally challenging decay signatures and low natural abundances of suitable candidate nuclei. In this article, we propose a new detector concept to overcome these challenges. It is based on the first-time combination of hybrid and opaque scintillation detector technology paired with novel light read-out techniques. This approach is particularly suitable detecting positrons (beta plus) signatures. We expect to discover two-neutrino double beta plus decay modes within 1 tonne-week exposure and are able to probe neutrinoless double beta plus decays at several orders of magnitude improved significance compared to current experimental limits., Comment: 23 pages, 9 figures, 3 tables

Hot subdwarfs (sdO/B) are the stripped helium cores of red giants formed by binary interactions. Close hot subdwarf binaries with massive white dwarf companions have been proposed as possible progenitors of thermonuclear supernovae type Ia (SN Ia). If the supernova is triggered by stable mass transfer from the helium star, the companion should survive the explosion and should be accelerated to high velocities. The hypervelocity star US 708 is regarded as the prototype for such an ejected companion. To find more of those objects we conducted an extensive spectroscopic survey. Candidates for such fast stars have been selected from the spectroscopic database of the Sloan Digital Sky Survey (SDSS) and several ground-based proper motion surveys. Follow-up spectroscopy has been obtained with several 4m- to 10m-class telescopes. Combining the results from quantitative spectroscopic analyses with space-based astrometry from \textit{Gaia} Early Data Release 3 (EDR3) we determined the atmospheric and kinematic parameters of 53 fast hot subdwarf stars. None of these stars is unbound to the Galaxy, although some have Galactic restframe velocities close to the Galactic escape velocity. 21 stars are apparently single objects, which crossed the Galactic disc within their lifetimes in the sdO/B stage and could be regarded as potential candidates for the SN Ia ejection scenario. However, the properties of the full sample are more consistent with a pure old Galactic halo population. We therefore conclude that the fast sdO/B stars we found are likely to be extreme halo stars., Comment: 22 pages, A&A accepted

Using novel metallic magnetic calorimeter detectors at the CRYRING@ESR, we recorded X-ray spectra of stored and electron cooled helium-like uranium (U$^{90+}$) with an unmatched spectral resolution of close to 90 eV. This allowed for an accurate determination of the energies of all four components of the K$\alpha$ transitions in U$^{90+}$. We find good agreement with state-of-the-art bound-state QED calculations for the strong-field regime. Our results do not support any systematic deviation between experiment and theory in helium-like systems, the presence of which was subject of intense debates in recent years.

Treating images as data has become increasingly popular in political science. While existing classifiers for images reach high levels of accuracy, it is difficult to systematically assess the visual features on which they base their classification. This paper presents a two-level classification method that addresses this transparency problem. At the first stage, an image segmenter detects the objects present in the image and a feature vector is created from those objects. In the second stage, this feature vector is used as input for standard machine learning classifiers to discriminate between images. We apply this method to a new dataset of more than 140,000 images to detect which ones display political protest. This analysis demonstrates three advantages to this paper's approach. First, identifying objects in images improves transparency by providing human-understandable labels for the objects shown on an image. Second, knowing these objects enables analysis of which distinguish protest images from non-protest ones. Third, comparing the importance of objects across countries reveals how protest behavior varies. These insights are not available using conventional computer vision classifiers and provide new opportunities for comparative research.

We present proper motion (PM) measurements for a sample of 23 massive star clusters within the Large Magellanic Cloud using multi-epoch data from the Hubble Space Telescope (HST). We combined archival data from the ACS/WFC and WFC3/UVIS instruments with observations from a dedicated HST programme, resulting in time baselines between 4.7 and 18.2 yr available for PM determinations. For bright well-measured stars, we achieved nominal PM precisions of 55 $\mu$as/yr down to 11 $\mu$as/yr . To demonstrate the potential and limitations of our PM data set, we analysed the cluster NGC 1850 and showcase a selection of different science applications. The precision of the PM measurements allows us to disentangle the kinematics of the various stellar populations that are present in the HST field. The cluster has a centre-of-mass motion that is different from the surrounding old field stars and also differs from the mean motion of a close-by group of very young stars. We determined the velocity dispersion of field stars to be 0.128 +/- 0.003 mas/yr (corresponding to 30.3 +/- 0.7 km/s). The velocity dispersion of the cluster inferred from the PM data set most probably overestimates the true value, suggesting that the precision of the measurements at this stage is not sufficient for a reliable analysis of the internal kinematics of extra-galactic star clusters. Finally, we exploit the PM-cleaned catalogue of likely cluster members to determine any radial segregation between fast and slowly-rotating stars, finding that the former are more centrally concentrated. With this paper, we also release the astro-photometric catalogues for each cluster., Comment: 27 pages, 15 figures. Accepted for publication in A&A. Astro-photometric catalogues are available at https://archive.stsci.edu/hlsp/hamsters

BL and UV Ceti are a nearby (2.7 pc) binary system with similar masses, spectral types, and rapid rotation rates, but very different magnetic activity. UV Ceti's much stronger large-scale magnetic field may cause this difference, highlighting key unanswered questions about dynamo processes in fully convective objects. Here we present multi-epoch characterization of the radio spectrum of UV Ceti spanning 1-105 GHz, exhibiting flared emission similar to coronal activity, auroral-like emission analogous to planetary magnetospheres, and slowly-varying persistent emission. Radio observations are a powerful means to probe the role that the large-scale magnetic field of UV Ceti has in non-thermal particle acceleration, because radio-frequency phenomena result from both the activity of small-scale field features as well as large-scale auroral current systems. We find temporal variability at all bands observed, and a hint of rotational modulation in the degree of circular polarization up to 40 GHz. The persistent component of the emission is fairly constant from 1-105 GHz, making optically thick emission or optically thin gyrosynchrotron from electrons with an isotropic pitch angle distribution unlikely. We discuss the possibility of emission mechanisms analogous to Jupiter's radiation belts., Comment: Accepted to ApJ

This paper introduces the Bosch street dataset (BSD), a novel multi-modal large-scale dataset aimed at promoting highly automated driving (HAD) and advanced driver-assistance systems (ADAS) research. Unlike existing datasets, BSD offers a unique integration of high-resolution imaging radar, lidar, and camera sensors, providing unprecedented 360-degree coverage to bridge the current gap in high-resolution radar data availability. Spanning urban, rural, and highway environments, BSD enables detailed exploration into radar-based object detection and sensor fusion techniques. The dataset is aimed at facilitating academic and research collaborations between Bosch and current and future partners. This aims to foster joint efforts in developing cutting-edge HAD and ADAS technologies. The paper describes the dataset's key attributes, including its scalability, radar resolution, and labeling methodology. Key offerings also include initial benchmarks for sensor modalities and a development kit tailored for extensive data analysis and performance evaluation, underscoring our commitment to contributing valuable resources to the HAD and ADAS research community.

We study the effects of anomalous thresholds on the non-local form factors describing the hadronization of the light-quark contribution to $B\to(P,V)\gamma^*$ transitions. Starting from a comprehensive discussion of anomalous thresholds in the triangle loop function for different mass configurations, we detail how the dispersion relation for $\pi\pi$ intermediate states is affected by contour deformations mandated by the anomalous branch points. Phenomenological estimates of the size of the anomalous contributions to the form factors are provided with couplings determined from measured branching fractions and Dalitz plot distributions. Our key finding is that anomalous effects are suppressed on the $\rho(770)$ resonance, while off-peak the effects can become as large as $\mathcal{O}(10\%)$ of the full (light-quark-loop induced) non-local form factors. We comment on future generalizations towards higher intermediate states and the charm loop, outlining how the dispersive framework established in this work could help improve the non-local form factors needed as input for a robust interpretation of $B\to (P,V)\ell^+\ell^-$ decays., Comment: 55 pages, 24 figures; journal version

More than 300 supermassive black holes have been detected at redshifts larger than 6, and they are abundant in the centers of local galaxies. Their formation mechanisms, however, are still rather unconstrained. A possible origin of these supermassive black holes could be through mergers in dense black hole clusters, forming as a result of mass segregation within nuclear star clusters in the center of galaxies. In this study, we present the first systematic investigation of the evolution of such black hole clusters where the effect of an external potential is taken into account. Such a potential could be the result of gas inflows into the central region, for example as a result of galaxy mergers. We show here that the efficiency for the formation of a massive central object is mostly regulated by the ratio of cluster velocity dispersion divided by the speed of light, potentially reaching efficiencies of 0.05-0.08 in realistic systems. Our results show that this scenario is potentially feasible and may provide seeds black hole of at least 10^3 solar masses. We conclude that the formation of seed black holes via this channel should be taken into account in statistical assessments of the black hole population., Comment: 15 pages, 11 figures, 1 table. Accepted for publication with A&A

Autonomous vehicles require a precise understanding of their environment to navigate safely. Reliable identification of unknown objects, especially those that are absent during training, such as wild animals, is critical due to their potential to cause serious accidents. Significant progress in semantic segmentation of anomalies has been driven by the availability of out-of-distribution (OOD) benchmarks. However, a comprehensive understanding of scene dynamics requires the segmentation of individual objects, and thus the segmentation of instances is essential. Development in this area has been lagging, largely due to the lack of dedicated benchmarks. To address this gap, we have extended the most commonly used anomaly segmentation benchmarks to include the instance segmentation task. Our evaluation of anomaly instance segmentation methods shows that this challenge remains an unsolved problem. The benchmark website and the competition page can be found at: https://vision.rwth-aachen.de/oodis ., Comment: Accepted at the VAND 2.0 Workshop at CVPR 2024. Project page: https://vision.rwth-aachen.de/oodis

The static and genuine structure of small rhodium and rhodium/tungsten nanoparticles on an alumina support can be imaged with atomic resolution even if single digit atom clusters are investigated. Low dose rate electron microscopy is key to the achievement and can generally be applied to investigate any similar material. In such conditions it becomes feasible to identify the chemical composition of nanocrystals from quantitative contrast analyses alone by counting atoms. The ability to fully characterize an unaltered, initial state of the objects allows targeting structural excitations or conformational changes induced by the electron beam itself. For the specific case of catalytic Rh:W particles we stimulate a tungsten-promoted size regulation mechanism in real time that is driven by Oswald ripening and can be understood by a strong binding of tungsten atoms to the oxygen atoms of the support, which builds up strain as the cluster sizes increase., Comment: 30 pages, 9 figures, 1 table

We introduce a novel generative model for the representation of joint probability distributions of a possibly large number of discrete random variables. The approach uses measure transport by randomized assignment flows on the statistical submanifold of factorizing distributions, which also enables to sample efficiently from the target distribution and to assess the likelihood of unseen data points. The embedding of the flow via the Segre map in the meta-simplex of all discrete joint distributions ensures that any target distribution can be represented in principle, whose complexity in practice only depends on the parametrization of the affinity function of the dynamical assignment flow system. Our model can be trained in a simulation-free manner without integration by conditional Riemannian flow matching, using the training data encoded as geodesics in closed-form with respect to the e-connection of information geometry. By projecting high-dimensional flow matching in the meta-simplex of joint distributions to the submanifold of factorizing distributions, our approach has strong motivation from first principles of modeling coupled discrete variables. Numerical experiments devoted to distributions of structured image labelings demonstrate the applicability to large-scale problems, which may include discrete distributions in other application areas. Performance measures show that our approach scales better with the increasing number of classes than recent related work.

Graph neural networks (GNNs) have emerged as powerful tools for processing relational data in applications. However, GNNs suffer from the problem of oversmoothing, the property that the features of all nodes exponentially converge to the same vector over layers, prohibiting the design of deep GNNs. In this work we study oversmoothing in graph convolutional networks (GCNs) by using their Gaussian process (GP) equivalence in the limit of infinitely many hidden features. By generalizing methods from conventional deep neural networks (DNNs), we can describe the distribution of features at the output layer of deep GCNs in terms of a GP: as expected, we find that typical parameter choices from the literature lead to oversmoothing. The theory, however, allows us to identify a new, nonoversmoothing phase: if the initial weights of the network have sufficiently large variance, GCNs do not oversmooth, and node features remain informative even at large depth. We demonstrate the validity of this prediction in finite-size GCNs by training a linear classifier on their output. Moreover, using the linearization of the GCN GP, we generalize the concept of propagation depth of information from DNNs to GCNs. This propagation depth diverges at the transition between the oversmoothing and non-oversmoothing phase. We test the predictions of our approach and find good agreement with finite-size GCNs. Initializing GCNs near the transition to the non-oversmoothing phase, we obtain networks which are both deep and expressive.

The prevailing approaches in Network Intrusion Detection Systems (NIDS) are often hampered by issues such as high resource consumption, significant computational demands, and poor interpretability. Furthermore, these systems generally struggle to identify novel, rapidly changing cyber threats. This paper delves into the potential of incorporating Neurosymbolic Artificial Intelligence (NSAI) into NIDS, combining deep learning's data-driven strengths with symbolic AI's logical reasoning to tackle the dynamic challenges in cybersecurity, which also includes detailed NSAI techniques introduction for cyber professionals to explore the potential strengths of NSAI in NIDS. The inclusion of NSAI in NIDS marks potential advancements in both the detection and interpretation of intricate network threats, benefiting from the robust pattern recognition of neural networks and the interpretive prowess of symbolic reasoning. By analyzing network traffic data types and machine learning architectures, we illustrate NSAI's distinctive capability to offer more profound insights into network behavior, thereby improving both detection performance and the adaptability of the system. This merging of technologies not only enhances the functionality of traditional NIDS but also sets the stage for future developments in building more resilient, interpretable, and dynamic defense mechanisms against advanced cyber threats. The continued progress in this area is poised to transform NIDS into a system that is both responsive to known threats and anticipatory of emerging, unseen ones.

For reconstructing high-fidelity human 3D models from monocular videos, it is crucial to maintain consistent large-scale body shapes along with finely matched subtle wrinkles. This paper explores the observation that the per-frame rendering results can be factorized into a pose-independent component and a corresponding pose-dependent equivalent to facilitate frame consistency. Pose adaptive textures can be further improved by restricting frequency bands of these two components. In detail, pose-independent outputs are expected to be low-frequency, while highfrequency information is linked to pose-dependent factors. We achieve a coherent preservation of both coarse body contours across the entire input video and finegrained texture features that are time variant with a dual-branch network with distinct frequency components. The first branch takes coordinates in canonical space as input, while the second branch additionally considers features outputted by the first branch and pose information of each frame. Our network integrates the information predicted by both branches and utilizes volume rendering to generate photo-realistic 3D human images. Through experiments, we demonstrate that our network surpasses the neural radiance fields (NeRF) based state-of-the-art methods in preserving high-frequency details and ensuring consistent body contours.

Cluster-based algorithm selection deals with selecting recommendation algorithms on clusters of users to obtain performance gains. No studies have been attempted for many combinations of clustering approaches and recommendation algorithms. We want to show that clustering users prior to algorithm selection increases the performance of recommendation algorithms. Our study covers eight datasets, four clustering approaches, and eight recommendation algorithms. We select the best performing recommendation algorithm for each cluster. Our work shows that cluster-based algorithm selection is an effective technique for optimizing recommendation algorithm performance. For five out of eight datasets, we report an increase in nDCG@10 between 19.28% (0.032) and 360.38% (0.191) compared to algorithm selection without prior clustering., Comment: 16 pages, 8 figures, 2 tables

One key challenge in backdoor attacks against large foundation models is the resource limits. Backdoor attacks usually require retraining the target model, which is impractical for very large foundation models. Existing backdoor attacks are mainly designed for supervised classifiers or small foundation models (e.g., BERT). None of these attacks has successfully compromised a very large foundation model, such as Llama-3-70B, especially with limited computational resources. In this paper, we propose TrojFM, a novel backdoor attack tailored for very large foundation models. Our primary technical contribution is the development of a novel backdoor injection method. This method forces a backdoored model to generate similar hidden representations for poisoned inputs regardless of their actual semantics. Our approach injects such backdoors by fine-tuning only a very small proportion of model parameters. This enables TrojFM to efficiently launch downstream task-agnostic backdoor attacks against very large foundation models under limited computational resources. Moreover, we optimize the fine-tuning process with our customized QLoRA technique, enabling launching our attack via only~\textit{one A100 GPU}. Furthermore, we design a new trigger injection method to ensure our attack stealthiness. Through extensive experiments, we first demonstrate that TrojFM can launch effective backdoor attacks against widely used large GPT-style models without jeopardizing their normal functionalities (and outperforming existing attacks on BERT-style models). Furthermore, we show that TrojFM is resilient to SOTA defenses and is insensitive to changes in key hyper-parameters. Finally, we conduct a resource analysis to quantify that our method can significantly save computational and memory costs compared to existing backdoor attacks.

This study investigates the anisotropic effects on a system of three qubits with chain and ring topology, described by the antiferromagnetic Heisenberg XXX model subjected to a homogeneous magnetic field. We explore the Stirling and Otto cycles and find that easy-axis anisotropy significantly enhances engine efficiency across all cases. At low temperatures, the ring configuration outperforms the chain on both work and efficiency during the Stirling cycle. Additionally, in both topologies, the Stirling cycle achieves Carnot efficiency with finite work at quantum critical points. In contrast, the quasistatic Otto engine also reaches Carnot efficiency at these points but yields no useful work. Notably, the Stirling cycle exhibits all thermal operational regimes engine, refrigerator, heater, and accelerator unlike the quasistatic Otto cycle, which functions only as an engine or refrigerator.

In this paper we investigate the chiral magnetic effect (CME): the generation of an electric current due to a homogeneous background magnetic field and a homogeneous chiral imbalance in QCD. We demonstrate that the leading coefficient describing the CME vanishes in equilibrium, both for free fermions as well as in full QCD. Our full QCD results are based on continuum extrapolated lattice simulations using dynamical staggered quarks with physical masses as well as quenched Wilson quarks. We show that it is crucial that a gauge invariant ultraviolet regularization is used to compute the CME and elaborate on why some of the existing in-equilibrium calculations of this effect gave a nonzero result. We stress that our findings imply the absence of a time-independent CME current flowing in equilibrium QCD, but do not concern the CME as an out-of-equilibrium, time-dependent effect., Comment: 30 pages, 9 figures

Application virtual machines provide strong isolation properties and are established in the context of software portability. Those opportunities make them interesting for scalable and secure IoT deployments. WebAssembly is an application virtual machine with origins in web browsers, that is getting rapidly adopted in other domains. The strong and steadily growing ecosystem makes WebAssembly an interesting candidate for Embedded Systems. This position paper discusses the usage of WebAssembly in Embedded Systems. After introducing the basic concepts of WebAssembly and existing runtime environments, we give an overview of the challenges for the efficient usage of WebAssembly in Embedded Systems. The paper concludes with a real world case study that demonstrates the viability, before giving an outlook on open issues and upcoming work.

It is now possible to estimate 3D human pose from monocular images with off-the-shelf 3D pose estimators. However, many practical applications require fine-grained absolute pose information for which multi-view cues and camera calibration are necessary. Such multi-view recordings are laborious because they require manual calibration, and are expensive when using dedicated hardware. Our goal is full automation, which includes temporal synchronization, as well as intrinsic and extrinsic camera calibration. This is done by using persons in the scene as the calibration objects. Existing methods either address only synchronization or calibration, assume one of the former as input, or have significant limitations. A common limitation is that they only consider single persons, which eases correspondence finding. We attain this generality by partitioning the high-dimensional time and calibration space into a cascade of subspaces and introduce tailored algorithms to optimize each efficiently and robustly. The outcome is an easy-to-use, flexible, and robust motion capture toolbox that we release to enable scientific applications, which we demonstrate on diverse multi-view benchmarks. Project website: https://github.com/jamestang1998/CasCalib., Comment: Accepted to the 18th IEEE International Conference on Automatic Face and Gesture Recognition

We study thermal freeze-out in the \textit{dark axion portal} - an axion coupling to the dark photon and photon - considering the dark photon as the lightest stable particle, thereby a dark matter candidate. We pay special attention to the coscattering regime, founding viable DM scenarios fulfilling the correct relic abundance. Furthermore, we explore the prospects of having the axion as a long-lived particle, possibly to be observed at the LHC and future detectors.

In the course of the energy transition, the expansion of generation and consumption will change, and many of these technologies, such as PV systems, electric cars and heat pumps, will influence the power flow, especially in the distribution grids. Scalable methods that can make decisions for each grid connection are needed to enable congestion-free grid operation in the distribution grids. This paper presents a novel end-to-end approach to resolving congestion in distribution grids with deep reinforcement learning. Our architecture learns to curtail power and set appropriate reactive power to determine a non-congested and, thus, feasible grid state. State-of-the-art methods such as the optimal power flow (OPF) demand high computational costs and detailed measurements of every bus in a grid. In contrast, the presented method enables decisions under sparse information with just some buses observable in the grid. Distribution grids are generally not yet fully digitized and observable, so this method can be used for decision-making on the majority of low-voltage grids. On a real low-voltage grid the approach resolves 100\% of violations in the voltage band and 98.8\% of asset overloads. The results show that decisions can also be made on real grids that guarantee sufficient quality for congestion-free grid operation.

This paper investigates the incidence of coherent emission in solar radio bursts, using a revised catalog of 3800 solar radio bursts observed by the Nobeyama Radio Polarimeters from 1988 to 2023. We focus on the 1.0 and 2.0 GHz data, where radio fluxes of order 10 billion Jansky have been observed. Previous work has suggested that these bursts are due to electron cyclotron maser (ECM) emission. In at least one well studied case, the bright emission at 1 GHz consists of narrowband spikes of millisecond duration. Coherent emission at 1 GHz can be distinguished from traditional incoherent gyrosynchrotron flare emission based on the radio spectrum: gyrosynchrotron emission at 1 GHz usually has a spectrum rising with frequency, so bursts in which 1 GHz is stronger than higher frequency measurements are unlikely to be incoherent gyrosynchrotron. Based on this criterion it is found that, for bursts exceeding 100 sfu, three-quarters of all bursts at 1 GHz and half of all 2 GHz bursts have a dominant coherent emission component, assumed to be ECM. The majority of the very bright bursts at 1 GHz are highly circularly polarized, consistent with a coherent emission mechanism, but not always 100% polarized. The frequency range from 1 to 2 GHz is heavily utilized for terrestrial applications, and these results are relevant for understanding the extreme flux levels that may impact such applications. Further, they provide a reference for comparison with the study of ECM emission from other stars and potentially exoplanets., Comment: 24 pages, 6 figures, in press

Purpose: Paranasal anomalies, frequently identified in routine radiological screenings, exhibit diverse morphological characteristics. Due to the diversity of anomalies, supervised learning methods require large labelled dataset exhibiting diverse anomaly morphology. Self-supervised learning (SSL) can be used to learn representations from unlabelled data. However, there are no SSL methods designed for the downstream task of classifying paranasal anomalies in the maxillary sinus (MS). Methods: Our approach uses a 3D Convolutional Autoencoder (CAE) trained in an unsupervised anomaly detection (UAD) framework. Initially, we train the 3D CAE to reduce reconstruction errors when reconstructing normal maxillary sinus (MS) image. Then, this CAE is applied to an unlabelled dataset to generate coarse anomaly locations by creating residual MS images. Following this, a 3D Convolutional Neural Network (CNN) reconstructs these residual images, which forms our SSL task. Lastly, we fine-tune the encoder part of the 3D CNN on a labelled dataset of normal and anomalous MS images. Results: The proposed SSL technique exhibits superior performance compared to existing generic self-supervised methods, especially in scenarios with limited annotated data. When trained on just 10% of the annotated dataset, our method achieves an Area Under the Precision-Recall Curve (AUPRC) of 0.79 for the downstream classification task. This performance surpasses other methods, with BYOL attaining an AUPRC of 0.75, SimSiam at 0.74, SimCLR at 0.73 and Masked Autoencoding using SparK at 0.75. Conclusion: A self-supervised learning approach that inherently focuses on localizing paranasal anomalies proves to be advantageous, particularly when the subsequent task involves differentiating normal from anomalous maxillary sinuses. Access our code at https://github.com/mtec-tuhh/self-supervised-paranasal-anomaly

We study the coscattering mechanism in a simple Higgs portal which add two real singlet scalars to the Standard Model. In this scenario, the lighter scalar is stabilized by a single $\mathcal{Z}_2$ symmetry and acts as the dark matter relic, whose freeze-out is driven by conversion processes. The heavier scalar becomes an unstable state which participate actively in the coscattering. We find viable parameter regions fulfilling the measured relic abundance, while evading direct detection and big-bang nucleosynthesis bounds. In addition, we discuss collider prospects for the heavier scalar as a long-lived particle at present and future detectors.

Interpretable machine learning has emerged as central in leveraging artificial intelligence within high-stakes domains such as healthcare, where understanding the rationale behind model predictions is as critical as achieving high predictive accuracy. In this context, feature selection assumes a pivotal role in enhancing model interpretability by identifying the most important input features in black-box models. While random forests are frequently used in biomedicine for their remarkable performance on tabular datasets, the accuracy gained from aggregating decision trees comes at the expense of interpretability. Consequently, feature selection for enhancing interpretability in random forests has been extensively explored in supervised settings. However, its investigation in the unsupervised regime remains notably limited. To address this gap, the study introduces novel methods to construct feature graphs from unsupervised random forests and feature selection strategies to derive effective feature combinations from these graphs. Feature graphs are constructed for the entire dataset as well as individual clusters leveraging the parent-child node splits within the trees, such that feature centrality captures their relevance to the clustering task, while edge weights reflect the discriminating power of feature pairs. Graph-based feature selection methods are extensively evaluated on synthetic and benchmark datasets both in terms of their ability to reduce dimensionality while improving clustering performance, as well as to enhance model interpretability. An application on omics data for disease subtyping identifies the top features for each cluster, showcasing the potential of the proposed approach to enhance interpretability in clustering analyses and its utility in a real-world biomedical application.

Single-molecule experiments provide insight into the motion (conformational dynamics) of individual protein molecules. Usually, a well-defined but coarse-grained intramolecular coordinate is measured and subsequently analysed with the help of Hidden Markov Models (HMMs) to deduce the kinetics of protein conformational changes. Such approaches rely on the assumption that the microscopic dynamics of the protein evolve according to a Markov-jump process on some network. However, the manifestation and extent of memory in the dynamics of the observable strongly depends on the chosen underlying Markov model, which is generally not known and therefore can lead to misinterpretations. Here, we combine extensive single-molecule plasmon ruler experiments on the heat shock protein Hsp90, computer simulations, and theory to infer and quantify memory in a model-free fashion. Our analysis is based on the bare definition of non-Markovian behaviour and does not require any underlying model. In the case of Hsp90 probed by a plasmon ruler, the Markov assumption is found to be clearly and conclusively violated on timescales up to roughly 50 s, which corresponds roughly to $\sim$50% of the inferred correlation time of the signal. The extent of memory is striking and reaches biologically relevant timescales. This implies that memory effects penetrate even the slowest observed motions. We provide clear and reproducible guidelines on how to test for the presence and duration of memory in experimental single-molecule data.

Supercontinua are broadband spectra that are essential to optical spectroscopy, sensing, imaging, and metrology. They are generated from ultrashort laser pulses through nonlinear frequency conversion in fibers, bulk media, and chip-integrated waveguides. For any generating platform, balancing the competing criteria of strong nonlinearity, transparency, and absence of multiphoton absorption is a key challenge. Here, we explore supercontinuum generation in integrated gallium nitride (GaN) waveguides, which combine a high Kerr-nonlinearity, mid-infrared transparency, and a large bandgap that prevents two- and three-photon absorption in the technologically important telecom C-band, where compact erbium-based pump lasers exist. Using this type of laser, we demonstrate tunable dispersive waves and gap-free spectra extending to almost 4 micron in wavelength, relevant to functional group chemical sensing. Additionally, leveraging the material's second-order nonlinearity, we implement on-chip f-2f interferometry to detect the pump laser's carrier-envelope offset frequency, which enables precision metrology. These results demonstrate the versatility of GaN-on-sapphire as a new platform for broadband nonlinear photonics.