Search

"Fluffy" hydrogenated amorphouscarbon(a-C:H)wassynthesizedusingadielectric barrier discharge plasma, driven by nanosec ond high voltage pulses at 1 kHz frequency in a helium-butane mixture. The a-C:H samples were characterized by scanning and transmission electron microscopy, laser-assisted and secondary ion mass spectrometry, and Raman and Fourier-transform infrared spectroscopy. We find that a-C:H samples exhibit infrared absorption features in good agreement with those observed for carbonaceous dust in IRAS 08572+3915 galaxy. We discuss their nano to microscale structure and derive their hydrogen to carbon (H/C) ratios from the results obtained by three distinct experimental characterization techniques. Relying on the average H/C value determined by mass spectrometry and Raman spectroscopy, we can then constrain the absorption strengths values to those best corresponding to our dust analogue, and calculate the H/C ratio from the infrared spectra. Altogether, we find that our dust analogue consists of a dominant hydrogen-rich aliphatic network, with small, isolated, aromatic regions. The a-C:H dust analogue was then irradiated with 3 MeV H+ and subsequently analyzed ex situ. Morphological and chemical changes, including the evolution of H/C, CH2/CH3, and sp2/sp3 ratios, were observed with increasing proton fluence, indicating dehydrogenation and graphitization. Proton bombardment shifted the initial location of a-C:H in the hydrocarbon ternary phase diagram toward the central region defined by IRAS 08572+3915 observations. The decay of the 3.4 {\mu}m band with proton fluence was used to calculate CH destruction cross-sections, results consistent with a direct effect of cosmic rays on the disappearance of the 3.4 {\mu}m band.

We acquired 450 {\mu}m and 850 {\mu}m dust continuum polarization observations toward the inner region of the Central Molecular Zone (CMZ) as part of the B-Fields In Star-Forming Region Observations (BISTRO) survey using the POL-2 polarimeter on the James Clerk Maxwell Telescope. These observations encompassed three dense structures: the 20 km s{^{-1}} cloud (20MC), 50 km s{^{-1}} cloud (50MC), and circumnuclear disk (CND). Our aim is to investigate the magnetic field morphology and strength in the inner region of the CMZ using polarized dust continuum and the Davis-Chandrasekhar-Fermi method. The magnetic field morphology is highly ordered in all three dense regions. The plane-of-sky magnetic field strengths are {\sim}1 mG for the 20MC and the 50MC, and {\sim}2 mG for the CND. We compare the energy contributions of turbulence, gravity, and thermal motion with that of the magnetic field using the plasma {\beta}, mass-to-flux ratio, and Alfv\'en Mach number. The outcomes reveal the magnetic field stands out as the predominant factor within the inner region of the CMZ. The dominance of the magnetic field may explain the low star-forming rate in the CMZ. We further investigate the dust grain alignment efficiency by exploring the relationship between polarization fraction and total intensity. The results suggest that dust grains are well aligned with the magnetic fields., Comment: Accepted for publication in ApJ. 31 pages, 21 figures (20 in main text, 1 in appendix), 1 appendix

In many industrial robotics applications, multiple robots are working in a shared workspace to complete a set of tasks as quickly as possible. Such settings can be treated as multi-modal multi-robot multi-goal path planning problems, where each robot has to reach an ordered sequence of goals. Existing approaches to this type of problem solve this using prioritization or assume synchronous completion of tasks, and are thus neither optimal nor complete. We formalize this problem as a single path planning problem and introduce a benchmark encompassing a diverse range of problem instances including scenarios with various robots, planning horizons, and collaborative tasks such as handovers. Along with the benchmark, we adapt an RRT* and a PRM* planner to serve as a baseline for the planning problems. Both planners work in the composite space of all robots and introduce the required changes to work in our setting. Unlike existing approaches, our planner and formulation is not restricted to discretized 2D workspaces, supports a changing environment, and works for heterogeneous robot teams over multiple modes with different constraints, and multiple goals. Videos and code for the benchmark and the planners is available at https://vhartman.github.io/mrmg-planning/., Comment: 8 pages, 8 figures

The sawtooth-free hybrid scenario has been achieved recently in ASDEX Upgrade (AUG) with applied non-inductive current sources and auxiliary heating [A. Burckhart et al 2023 Nucl. Fusion 63 126056]. Control experiments in AUG suggest that the self-regulating magnetic flux pumping mechanism, characterized by anomalous current redistribution, is responsible for clamping the central safety factor (q_0) close to unity, thereby preventing the sawtooth onset. This work presents a numerical and theoretical investigation of flux pumping in the AUG hybrid scenario based on the two-temperature, visco-resistive, full magnetohydrodynamic (MHD) model with the JOREK code. To quantitatively model the flux pumping, we choose realistic parameters, plasma configurations, and source terms based on AUG experiments. During the initial saturation stage of the unstable 1/1 quasi-interchange mode (on millisecond timescales), q_0 exhibits fast under-damped oscillation and reaches a value closer to unity, which is attributed to the self-regulation of core plasma and the fast dynamo effect on the order of V/m. On the longer resistive diffusion timescale of seconds, the slow negative dynamo effect on the order of mV/m induced by the 1/1 MHD instability plays an effective role in flux pumping, which provides quantitative agreement with experimental observations for the first time. The final saturated 1/1 MHD instability exhibits features of the quasi-interchange mode and tearing mode, and the associated convective plasma flow velocity is a few m/s. The toroidal negative electric field from the slow dynamo dominantly offsets the positive current drive and continuously redistributes the current density and pressure. As a result, q_0 is maintained close to unity due to the low-shear profiles of current density and pressure in the plasma core, and the system enters a sawtooth-free and quasi-stationary helical state.

When manipulating objects in the real world, we need reactive feedback policies that take into account sensor information to inform decisions. This study aims to determine how different encoders can be used in a reinforcement learning (RL) framework to interpret the spatial environment in the local surroundings of a robot arm. Our investigation focuses on comparing real-world vision with 3D scene inputs, exploring new architectures in the process. We built on the SERL framework, providing us with a sample efficient and stable RL foundation we could build upon, while keeping training times minimal. The results of this study indicate that spatial information helps to significantly outperform the visual counterpart, tested on a box picking task with a vacuum gripper. The code and videos of the evaluations are available at https://github.com/nisutte/voxel-serl., Comment: 8 pager, 5 Figures, 5 Tables

In Precision Livestock Farming, detecting deviations from optimal or baseline values - i.e. anomalies in time series - is essential to allow undertaking corrective actions rapidly. Here we aim at detecting anomalies in 24h time series of cow activity, with a view to detect cases of disease or oestrus. Deviations must be distinguished from noise which can be very high in case of biological data. It is also important to detect the anomaly early, e.g. before a farmer would notice it visually. Here, we investigate the benefit of using wavelet transforms to denoise data and we assess the performance of an anomaly detection algorithm considering the timing of the detection. We developed features based on the comparisons between the wavelet transforms of the mean of the time series and the wavelet transforms of individual time series instances. We hypothesized that these features contribute to the detection of anomalies in periodic time series using a feature-based algorithm. We tested this hypothesis with two datasets representing cow activity, which typically follows a daily pattern but can deviate due to specific physiological or pathological conditions. We applied features derived from wavelet transform as well as statistical features in an Isolation Forest algorithm. We measured the distance of detection between the days annotated abnormal by animal caretakers days and the days predicted abnormal by the algorithm. The results show that wavelet-based features are among the features most contributing to anomaly detection. They also show that detections are close to the annotated days, and often precede it. In conclusion, using wavelet transforms on time series of cow activity data helps to detect anomalies related to specific cow states. The detection is often obtained on days that precede the day annotated by caretakers, which offer possibility to take corrective actions at an early stage., Comment: 17 pages, 8 figures, 4 tables, 1 algorithm

The Minkowski vacuum in an accelerated frame behaves like a fluid that has not only a finite temperature due to the Unruh effect, but also a finite shear viscosity. Moreover, the ratio of this viscosity to the entropy density exactly satisfies the Kovtun-Son-Starinets (KSS) bound, inspired by the string theory $ \eta/s=1/4\pi $. The origin of this viscosity is purely kinematical and is believed to be related to entanglement introduced by the Rindler horizon. We directly calculate the viscosity, entropy density, and their ratio for massless fields with spins 1/2 and 1. We show that locally the ratio of viscosity to entropy density can be below the limiting value $ 1/4\pi $ at distances of the order of the thickness of the membrane corresponding to the stretched horizon, and is described by the universal function for different spins. In particular, on the membrane surface $ \eta/s=1/8\pi $., Comment: 21 pages, 4 figures

Domain shift presents a significant challenge in applying Deep Learning to the segmentation of 3D medical images from sources like Magnetic Resonance Imaging (MRI) and Computed Tomography (CT). Although numerous Domain Adaptation methods have been developed to address this issue, they are often evaluated under impractical data shift scenarios. Specifically, the medical imaging datasets used are often either private, too small for robust training and evaluation, or limited to single or synthetic tasks. To overcome these limitations, we introduce a M3DA /"mEd@/ benchmark comprising four publicly available, multiclass segmentation datasets. We have designed eight domain pairs featuring diverse and practically relevant distribution shifts. These include inter-modality shifts between MRI and CT and intra-modality shifts among various MRI acquisition parameters, different CT radiation doses, and presence/absence of contrast enhancement in images. Within the proposed benchmark, we evaluate more than ten existing domain adaptation methods. Our results show that none of them can consistently close the performance gap between the domains. For instance, the most effective method reduces the performance gap by about 62% across the tasks. This highlights the need for developing novel domain adaptation algorithms to enhance the robustness and scalability of deep learning models in medical imaging. We made our M3DA benchmark publicly available: https://github.com/BorisShirokikh/M3DA., Comment: 17 pages,7 figures,11 tables

In this work we propose a satellite specific Neural Radiance Fields (NeRF) model capable to obtain a three-dimensional semantic representation (neural semantic field) of the scene. The model derives the output from a set of multi-date satellite images with corresponding pixel-wise semantic labels. We demonstrate the robustness of our approach and its capability to improve noisy input labels. We enhance the color prediction by utilizing the semantic information to address temporal image inconsistencies caused by non-stationary categories such as vehicles. To facilitate further research in this domain, we present a dataset comprising manually generated labels for popular multi-view satellite images. Our code and dataset are available at https://github.com/wagnva/semantic-nerf-for-satellite-data., Comment: Accepted at the CV4EO Workshop at WACV 2025

This work-in-progress paper presents a novel approach to detecting sponsored advertisement segments in YouTube videos and comparing the advertisement with the main content. Our methodology involves the collection of 421 auto-generated and manual transcripts which are then fed into a prompt-engineered GPT-4o for ad detection, a KeyBERT for keyword extraction, and another iteration of ChatGPT for category identification. The results revealed a significant prevalence of product-related ads across various educational topics, with ad categories refined using GPT-4o into succinct 9 content and 4 advertisement categories. This approach provides a scalable and efficient alternative to traditional ad detection methods while offering new insights into the types and relevance of ads embedded within educational content. This study highlights the potential of LLMs in transforming ad detection processes and improving our understanding of advertisement strategies in digital media., Comment: 6 pages, 4 figures, accepted and presented in the 10th IEEE International Conference on Sustainable Technology and Engineering

The rapid development of RISC-V instruction set architecture presents new opportunities and challenges for software developers. Is it sufficient to simply recompile high-performance software optimized for x86-64 onto RISC-V CPUs? Are current compilers capable of effectively optimizing C and C++ codes or is it necessary to use intrinsics or assembler? Can we analyze and improve performance without well-developed profiling tools? Do standard optimization techniques work? Are there specific RISC-V features that need to be considered? These and other questions require careful consideration. In this paper, we present our experience optimizing four BLAS algorithms for band matrix operations on RISC-V processors. We demonstrate how RISC-V-optimized implementations of OpenBLAS algorithms can be significantly accelerated through improved vectorization of computationally intensive loops. Experiments on Lichee Pi 4A and Banana Pi BPI-F3 devices using RVV 0.7.1 and RVV 1.0 vector instruction sets respectively, show speedups of 1.5x to 10x depending on the operation compared to the OpenBLAS baseline. In particular, the successful use of vector register grouping with RVV can lead to significant performance improvements.

In a seminal work, K. Segerberg introduced a deontic logic called DAL to investigate normative reasoning over actions. DAL marked the beginning of a new area of research in Deontic Logic by shifting the focus from deontic operators on propositions to deontic operators on actions. In this work, we revisit DAL and provide a complete algebraization for it. In our algebraization we introduce deontic action algebras -- algebraic structures consisting of a Boolean algebra for interpreting actions, a Boolean algebra for interpreting formulas, and two mappings from one Boolean algebra to the other interpreting the deontic concepts of permission and prohibition. We elaborate on how the framework underpinning deontic action algebras enables the derivation of different deontic action logics by removing or imposing additional conditions over either of the Boolean algebras. We leverage this flexibility to demonstrate how we can capture in this framework several logics in the DAL family. Furthermore, we introduce four variations of DAL by: (a) enriching the algebra of formulas with propositions on states, (b) adopting a Heyting algebra for state propositions, (c) adopting a Heyting algebra for actions, and (d) adopting Heyting algebras for both. We illustrate these new deontic action logics with examples and establish their algebraic completeness., Comment: This paper has been submitted to the *** DaL\'i - Dynamic Logic: New Trends and Applications *** Special issue of the Journal of Logic, Language and Information

Time domain simulations are crucial for analyzing transient behavior and broadband responses in electromagnetic problems. However, conventional numerical methods such as finite element method in time domain (FEMTD) and finite difference time domain, can be computationally demanding due to their high-dimensional nature, making large-scale simulations impractical for design optimization and real-time analysis. This paper introduces TA-ROMTD, a time-adaptive reduced order model (ROM) for FEMTD simulations that significantly reduces computational costs while maintaining accuracy. The method alternates between FEMTD and a reduced order model in time domain (ROMTD), using an error estimator to detect when the ROMTD solution loses accuracy and switching back to FEMTD to update the ROM with new data. Thus, TA-ROMTD does not require prior knowledge of the problem, as the ROM is constructed on the fly using FEMTD data. A key feature of this approach is the use of a coarse time step during FEMTD time intervals, capturing essential system dynamics while minimizing computational overhead. By reducing the number of degrees of freedom, this method enables efficient electromagnetic simulations, making it a powerful tool for antenna and microwave circuit design. The efficiency of the TA-ROMTD strategy is demonstrated through numerical examples, including an antipodal Vivaldi antenna, a dielectric resonator antenna, a fully metallic dual-polarization frequency selective surface, and a mushroom-type electromagnetic bandgap structure. These cases show the capability of the proposed approach to achieve accurate solutions while significantly reducing the computation time from tens of hours to minutes.

People worldwide use language in subtle and complex ways to express emotions. While emotion recognition -- an umbrella term for several NLP tasks -- significantly impacts different applications in NLP and other fields, most work in the area is focused on high-resource languages. Therefore, this has led to major disparities in research and proposed solutions, especially for low-resource languages that suffer from the lack of high-quality datasets. In this paper, we present BRIGHTER -- a collection of multilabeled emotion-annotated datasets in 28 different languages. BRIGHTER covers predominantly low-resource languages from Africa, Asia, Eastern Europe, and Latin America, with instances from various domains annotated by fluent speakers. We describe the data collection and annotation processes and the challenges of building these datasets. Then, we report different experimental results for monolingual and crosslingual multi-label emotion identification, as well as intensity-level emotion recognition. We investigate results with and without using LLMs and analyse the large variability in performance across languages and text domains. We show that BRIGHTER datasets are a step towards bridging the gap in text-based emotion recognition and discuss their impact and utility., Comment: 20 pages, under review

This work establishes the technical fundamentals of a well-tuned Customizable Contraction Hierarchies (CCH) implementation that is simple and elegant. We give a detailed overview of the state of the art of CCH, review recent advances on CCH and show how to combine them. Additionally, we propose further refinements that improve the performance of CCH. An extensive evaluation confirms that a CCH framework is not only comprehensive in supported features but also competitive in performance to both Contraction Hierarchies (CH) and Customizable Route Planning (CRP).

Finding a proper solution concept for the multi-dimensional barotropic compressible Euler equations and related systems is still an unsolved problem. As revealed by convex integration, the classical notion of an admissible weak solutions (also known as weak entropy solutions) does not lead to uniqueness and allows for solutions which do not seem to be physical. For this reason, people have studied additional criteria in view of their ability to rule out the counterintuitive solutions generated by convex integration. Recently, in [H.~Gimperlein, M.~Grinfeld, R.~J.~Knops and M.~Slemrod: The least action admissibility principle, arXiv: 2409.07191 (2024)] it was suggested that the least action admissibility principle serves as the desired selection criterion. In this paper, however, we show that the least action admissibility principle rules out the solution which is intuitively the physically relevant one. Consequently, one either has to reconsider one's intuition, or the least action admissibility principle must be discarded., Comment: 15 pages, 2 figures

One of the instruments aboard the Sunrise III mission, the Tunable Magnetograph (TuMag), is a tunable imaging spectropolarimeter in visible wavelengths. It is designed to probe the vector magnetic field and the line-of-sight velocity of the photosphere and the lower chromosphere. The quasi-simultaneous observation of two spectral lines provides excellent diagnostic measurements of the magnetic and dynamic coupling in these layers. The key technologies employed for TuMag are an LCVR-based polarimeter and a solid, LiNbO3 Fabry-P\'erot etalon as a spectrometer. However, it also incorporates several innovative features, such as home-made high-sensitivity scientific cameras and a double filter wheel. TuMag can sequentially observe any two out of the three spectral lines of Fe I at 525.02 and 525.06 nm and of Mg I at 517.3 nm. Laboratory measurements have demonstrated outstanding performance, including a wavefront root-mean-square error better than {\lambda}/13 for image quality, a full-width-at-half-maximum of 8.7 pm for the filtergraph transmission profile, and polarimetric efficiencies > 0.54. Here we report on the concept, design, calibration, and integration phases of the instrument, as well as on the data reduction pipeline., Comment: Contains 58 pages and 25 figures; to be published in Solar Physics Topical Collection "The Sunrise III Solar Observatory" (https://link.springer.com/collections/jegdciedig)

Large language models (LLMs) are helping millions of users write texts about diverse issues, and in doing so expose users to different ideas and perspectives. This creates concerns about issue bias, where an LLM tends to present just one perspective on a given issue, which in turn may influence how users think about this issue. So far, it has not been possible to measure which issue biases LLMs actually manifest in real user interactions, making it difficult to address the risks from biased LLMs. Therefore, we create IssueBench: a set of 2.49m realistic prompts for measuring issue bias in LLM writing assistance, which we construct based on 3.9k templates (e.g. "write a blog about") and 212 political issues (e.g. "AI regulation") from real user interactions. Using IssueBench, we show that issue biases are common and persistent in state-of-the-art LLMs. We also show that biases are remarkably similar across models, and that all models align more with US Democrat than Republican voter opinion on a subset of issues. IssueBench can easily be adapted to include other issues, templates, or tasks. By enabling robust and realistic measurement, we hope that IssueBench can bring a new quality of evidence to ongoing discussions about LLM biases and how to address them., Comment: under review

We present MatSwap, a method to transfer materials to designated surfaces in an image photorealistically. Such a task is non-trivial due to the large entanglement of material appearance, geometry, and lighting in a photograph. In the literature, material editing methods typically rely on either cumbersome text engineering or extensive manual annotations requiring artist knowledge and 3D scene properties that are impractical to obtain. In contrast, we propose to directly learn the relationship between the input material -- as observed on a flat surface -- and its appearance within the scene, without the need for explicit UV mapping. To achieve this, we rely on a custom light- and geometry-aware diffusion model. We fine-tune a large-scale pre-trained text-to-image model for material transfer using our synthetic dataset, preserving its strong priors to ensure effective generalization to real images. As a result, our method seamlessly integrates a desired material into the target location in the photograph while retaining the identity of the scene. We evaluate our method on synthetic and real images and show that it compares favorably to recent work both qualitatively and quantitatively. We will release our code and data upon publication.

In July 2024, Sunrise completed its third successful science flight. The Sunrise III observatory had been upgraded significantly after the two previous successful flights in 2009 and 2013. Three completely new instruments focus on the small-scale physical processes and their complex interaction from the deepest observable layers in the photosphere up to chromospheric heights. Previously poorly explored spectral regions and lines are exploited to paint a three-dimensional picture of the solar atmosphere with unprecedented completeness and level of detail. The full polarimetric information is captured by all three instruments to reveal the interaction between the magnetic fields and the hydrodynamic processes. Two slit-based spectropolarimeters, the Sunrise UV Spectropolarimeter and Imager (SUSI) and the Sunrise Chromospheric Infrared spectro-Polarimeter (SCIP), focus on the near-ultraviolet and the near-infrared regions respectively, and the imaging spectropolarimeter Tunable Magnetograph (TuMag) simultaneously obtains maps of the full field-of-view of $46 \times 46$ Mm$^2$ in the photosphere and the chromosphere in the visible. The instruments are operated in an orchestrated mode, benefiting from a new Image Stabilization and Light Distribution unit (ISLiD), with the Correlating Wavefront Sensor (CWS) providing the autofocus control and an image stability with a root-mean-square value smaller than 0.005''. A new gondola was constructed to significantly improve the telescope pointing stability, required to achieve uninterrupted observations over many hours. Sunrise III was launched successfully on July 10, 2024, from the Esrange Space Center near Kiruna (Sweden). It reached the landing site between the Mackenzie River and the Great Bear Lake in Canada after a flight duration of 6.5 days. In this paper, we give an overview of the Sunrise III observatory and its instruments., Comment: 67 pages, 25 figures; to be published in Solar Physics Topical Collection "The Sunrise III Solar Observatory" (https://link.springer.com/collections/jegdciedig)

Neural networks are fragile when confronted with data that significantly deviates from their training distribution. This is true in particular for simulation-based inference methods, such as neural amortized Bayesian inference (ABI), where models trained on simulated data are deployed on noisy real-world observations. Recent robust approaches employ unsupervised domain adaptation (UDA) to match the embedding spaces of simulated and observed data. However, the lack of comprehensive evaluations across different domain mismatches raises concerns about the reliability in high-stakes applications. We address this gap by systematically testing UDA approaches across a wide range of misspecification scenarios in both a controlled and a high-dimensional benchmark. We demonstrate that aligning summary spaces between domains effectively mitigates the impact of unmodeled phenomena or noise. However, the same alignment mechanism can lead to failures under prior misspecifications - a critical finding with practical consequences. Our results underscore the need for careful consideration of misspecification types when using UDA techniques to increase the robustness of ABI in practice.

We study the ultraviolet problem for models of a finite-dimensional quantum mechanical system linearly coupled to a bosonic quantum field, such as the (many-)spin boson model or its rotating-wave approximation. If the state change of the system upon emission or absorption of a boson is either given by a normal matrix or by a 2-nilpotent one, which is the case for the previously named examples, we prove an optimal renormalization result. We complement it, by proving the norm resolvent convergence of appropriately regularized models to the renormalized one. Our method consists of a dressing transformation argument in the normal case and an appropriate interior boundary condition for the 2-nilpotent case., Comment: 18 pages

We present a new variational formulation for Viscous and resistive Hall Magnetohydrodynamic. We first find a variational principle for ideal HMHD by applying the physical assumptions leading to HMHD at the lagrangian level, and then we add the viscous and resistive terms by the means of constrained variations. We also provide a metriplectic reformulation of our formulation, based on two canonical Lie-Poisson brackets for the ideal part and metric 4-brackets for the dissipative part., Comment: 21 pages

We study the sphere partition function of 3d $\mathcal{N}=2$ holographic SCFTs arising on the worldvolume of $N$ coincident M2-branes in the presence of squashing and real mass deformations. We argue that the all-order large $N$ perturbative expansion of the partition function resums into a compact expression in terms of an Airy function. We provide ample evidence for this conjecture using numerical methods, saddle point analysis in the 't Hooft limit, and the relation of the squashed sphere partition function to the Cardy-like limit of the superconformal index. We also discuss the relation of our results to topological string theory and the statistical mechanics of Fermi gases. Our conjecture has a bearing on the AdS/CFT correspondence and the structure of perturbative M-theory corrections to 11d supergravity which we discuss., Comment: v1: 64 pages

Recently, optically active spin defects embedded in van der Waals (vdW) crystals have emerged as a transformative quantum sensing platform to explore cutting-edge materials science and quantum physics. Taking advantage of excellent solid-state integrability, this new class of spin defects can be arranged in controllable nanoscale proximity of target materials in vdW heterostructures, showing great promise for improving spatial resolution and field sensitivity of current sensing technologies. Building on this state-of-the-art measurement platform, here we report hexagonal boron nitride-based quantum imaging of field-free deterministic magnetic switching of room-temperature two-dimensional magnet Fe3GaTe2 in an all-vdW spin-orbit torque (SOT) system. By visualizing SOT-driven variations of nanoscale Fe3GaTe2 magnetic stray field profile under different conditions, we have revealed how the observed magnetic switching evolves from deterministic to indeterministic behavior due to the interplay between out-of-plane spins, in-plane spins and Joule heating. This understanding, which is otherwise difficult to access by conventional transport measurements, offers valuable insights on material design, testing, and evaluation of next-generation vdW spintronic devices., Comment: 13 pages, 4 figures

The field of high redshift galaxy formation has been revolutionised by JWST, which is yielding unprecedented insights on galaxy assembly at early times. Our key aim is to study the physical mechanisms that can explain the unexpected abundance of bright galaxies at $z \geq 11$, as well as their metal enrichment and spectral properties. We also use recent data to determine the key sources of reionisation. To do so, we implement cold gas fractions and star formation efficiencies derived from the SPHINX20 high-resolution radiation-hydrodynamics simulation into DELPHI, a semi-analytic model that tracks the assembly of dark matter halos and their baryonic components from $z \sim 4.5-40$. In addition, we explore two different methodologies to boost galaxy luminosities at $z \geq 11$: a stellar initial mass function (IMF) that becomes increasingly top-heavy with decreasing metallicity and increasing redshift (eIMF model), and star formation efficiencies that increase with increasing redshift (eSFE model). Our key findings are: (i) both the eIMF and eSFE models can explain the abundance of bright galaxies at $z \geq 11$; (ii) dust attenuation plays an important role for the bright-end of the UV LF at $z \leq 11$; (iii) the mass-metallicity relation is in place as early as $z \sim 17$ in all models although its slope is model-dependent; (iv) within the spread of both models and observations, all of our models are in good agreement with current estimates of $\beta$ slopes at $z \sim 5-17$ and Balmer break strengths at $z \sim 6-10$; (v) in the eIMF model, galaxies at $z\geq12$ or with $\rm{M_{UV}}\geq-18$ show values of $\xi_{\rm{ion}} \sim 10^{25.55}~{\rm [Hz~erg^{-1}]}$, twice larger than in other models; (vi) star formation in galaxies below $10^{9}\rm{M_{\odot}}$ is the key driver of reionisation, providing the bulk ($\sim 85\%$) of ionising photons down to its midpoint at $z \sim 7$.

Diffusion models generate high-quality synthetic data. They operate by defining a continuous-time forward process which gradually adds Gaussian noise to data until fully corrupted. The corresponding reverse process progressively "denoises" a Gaussian sample into a sample from the data distribution. However, generating high-quality outputs requires many discretization steps to obtain a faithful approximation of the reverse process. This is expensive and has motivated the development of many acceleration methods. We propose to accomplish sample generation by learning the posterior {\em distribution} of clean data samples given their noisy versions, instead of only the mean of this distribution. This allows us to sample from the probability transitions of the reverse process on a coarse time scale, significantly accelerating inference with minimal degradation of the quality of the output. This is accomplished by replacing the standard regression loss used to estimate conditional means with a scoring rule. We validate our method on image and robot trajectory generation, where we consistently outperform standard diffusion models at few discretization steps.

We introduce a framework for describing the real-time dynamics of quantum impurity models out of equilibrium which is based on the influence matrix approach. By replacing the dynamical map of a large fermionic quantum environment with an effective semi-group influence matrix (SGIM) which acts on a reduced auxiliary space, we overcome the limitations of previous proposals, achieving high accuracy at long evolution times. This SGIM corresponds to a uniform matrix-product state representation of the influence matrix and can be obtained by an efficient algorithm presented in this paper. We benchmark this approach by computing the spectral function of the single impurity Anderson model with high resolution. Further, the spectrum of the effective dynamical map allows us to obtain relaxation rates of the impurity towards equilibrium following a quantum quench. Finally, for a quantum impurity model with on-site two-fermion loss, we compute the spectral function and confirm the emergence of Kondo physics at large loss rates., Comment: 6+3 pages, 4 figures

Probabilistic neurosymbolic learning seeks to integrate neural networks with symbolic programming. Many state-of-the-art systems rely on a reduction to the Probabilistic Weighted Model Counting Problem (PWMC), which requires computing a Boolean formula called the logical provenance.However, PWMC is \\#P-hard, and the number of clauses in the logical provenance formula can grow exponentially, creating a major bottleneck that significantly limits the applicability of PNL solutions in practice.We propose a new approach centered around an exact algorithm DPNL, that enables bypassing the computation of the logical provenance.The DPNL approach relies on the principles of an oracle and a recursive DPLL-like decomposition in order to guide and speed up logical inference.Furthermore, we show that this approach can be adapted for approximate reasoning with $\epsilon$ or $(\epsilon, \delta)$ guarantees, called ApproxDPNL.Experiments show significant performance gains.DPNL enables scaling exact inference further, resulting in more accurate models.Further, ApproxDPNL shows potential for advancing the scalability of neurosymbolic programming by incorporating approximations even further, while simultaneously ensuring guarantees for the reasoning process.

Quantifying uncertainty in neural networks is a highly relevant problem which is essential to many applications. The two predominant paradigms to tackle this task are Bayesian neural networks (BNNs) and deep ensembles. Despite some similarities between these two approaches, they are typically surmised to lack a formal connection and are thus understood as fundamentally different. BNNs are often touted as more principled due to their reliance on the Bayesian paradigm, whereas ensembles are perceived as more ad-hoc; yet, deep ensembles tend to empirically outperform BNNs, with no satisfying explanation as to why this is the case. In this work we bridge this gap by showing that deep ensembles perform exact Bayesian averaging with a posterior obtained with an implicitly learned data-dependent prior. In other words deep ensembles are Bayesian, or more specifically, they implement an empirical Bayes procedure wherein the prior is learned from the data. This perspective offers two main benefits: (i) it theoretically justifies deep ensembles and thus provides an explanation for their strong empirical performance; and (ii) inspection of the learned prior reveals it is given by a mixture of point masses -- the use of such a strong prior helps elucidate observed phenomena about ensembles. Overall, our work delivers a newfound understanding of deep ensembles which is not only of interest in it of itself, but which is also likely to generate future insights that drive empirical improvements for these models.

Y$_2$Co$_3$ is a newly discovered antiferromagnetic (AFM) compound with distorted kagome layers. Previous investigations via bulk magnetization measurements suggested a complex noncollinear magnetic behavior, with magnetic moments primarily anti-aligned along the $b$ axis and some canting towards the $ac$ plane. In this study, we report the magnetic structure of Y$_2$Co$_3$ to be an A-type AFM structure with ferromagnetic (FM) interactions within the distorted kagome plane and an interplane antiferromagnetic interaction, as determined by single-crystal neutron diffraction. The magnetic moments align along the $b$ axis, with minimal canting towards the $c$ axis, at odds with the previous interpretation of bulk magnetization measurements. The magnetic moments on the two distinct Co sites are [0, -0.68(3), 0] $\mu_B$ and [0, 1.25(4), 0.07(1)] $\mu_B$. We attribute the previously reported "noncollinear" behavior to the considerable temperature dependence of itinerant AFM exchange interactions, induced by thermal contraction along the $b$ axis. Additionally, our examination of lattice constants through pressure studies reveals compensating effects on FM and AFM interactions, resulting in negligible pressure dependence of $T_\textrm{N}$.

Cryo-electron tomography (cryo-ET) enables 3D visualization of cellular environments. Accurate reconstruction of high-resolution volumes is complicated by the very low signal-to-noise ratio and a restricted range of sample tilts, creating a missing wedge of Fourier information. Recent self-supervised deep learning approaches, which post-process initial reconstructions done by filtered backprojection (FBP), have significantly improved reconstruction quality, but they are computationally expensive, demand large memory, and require retraining for each new dataset. End-to-end supervised learning is an appealing alternative but is impeded by the lack of ground truth and the large memory demands of high-resolution volumetric data. Training on synthetic data often leads to overfitting and poor generalization to real data, and, to date, no general end-to-end deep learning reconstructors exist for cryo-ET. In this work, we introduce CryoLithe, a local, memory-efficient reconstruction network that directly estimates the volume from an aligned tilt-series, overcoming the suboptimal FBP. We demonstrate that leveraging transform-domain locality makes our network robust to distribution shifts, enabling effective supervised training and giving excellent results on real data -- without retraining or fine-tuning.

Neurons primarily communicate through the emission of action potentials, or spikes. To generate a spike, a neuron's membrane potential must cross a defined threshold. Does this spiking mechanism inherently prevent neurons from transmitting their subthreshold membrane potential fluctuations to other neurons? We prove that, in theory, it does not. The subthreshold membrane potential fluctuations of a presynaptic population of spiking neurons can be perfectly transmitted to a downstream population of neurons. Mathematically, this surprising result is an example of concentration phenomenon in high dimensions., Comment: 15 pages, 1 figure

Improving the efficiency of photothermal (PT) therapies and photoacoustic (PA) imaging at the microscopic scale hinges on developing multiphoton-absorbing photothermal molecules or contrast agents that operate in the near-infrared (NIR) range. These advanced agents or molecules will enable excitation with NIR lasers, in an improved transparency range for biological tissues, while enabling minute, highly localized spatial control of the excitation area. However, progress in this field requires innovative experimental techniques to characterize photothermal and photoacoustic effects under multiphoton excitation. In this article, we showcase a study of a model organometallic molecular compound excited via two-photon absorption (2PA) using femtosecond laser pulses. Based on a time-domain Brillouin scattering technique, well adapted for investigating ultrafast nonlinear optical absorption processes in ultrathin films on substrates, we determine the effective nonlinear absorption coefficients of the compound directly linked to PT/PA. Our findings provide a practical approach for exploring and optimizing nonlinear PT/PA absorbers and contrast agents.

Summary: VTX is a molecular visualization software capable to handle most molecular structures and dynamics trajectories file formats. It features a real-time high-performance molecular graphics engine, based on modern OpenGL, optimized for the visualization of massive molecular systems and molecular dynamics trajectories. VTX includes multiple interactive camera and user interaction features, notably free-fly navigation and a fully modular graphical user interface designed for increased usability. It allows the production of high-resolution images for presentations and posters with custom background. VTX design is focused on performance and usability for research, teaching and educative purposes. Availability and implementation: VTX is open source and free for non commercial use. Builds for Windows and Ubuntu Linux are available at http://vtx.drugdesign.fr. The source code is available at https://github.com/VTX-Molecular-Visualization, Comment: 5 pages, 1 figure, download VTX at http://vtx.drugdesign.fr, follow development at https://github.com/VTX-Molecular-Visualization

Systemic drug administration often causes off-target effects limiting the efficacy of advanced therapies. Targeted drug delivery approaches increase local drug concentrations at the diseased site while minimizing systemic drug exposure. We present a magnetically guided microrobotic drug delivery system capable of precise navigation under physiological conditions. This platform integrates a clinical electromagnetic navigation system, a custom-designed release catheter, and a dissolvable capsule for accurate therapeutic delivery. In vitro tests showed precise navigation in human vasculature models, and in vivo experiments confirmed tracking under fluoroscopy and successful navigation in large animal models. The microrobot balances magnetic material concentration, contrast agent loading, and therapeutic drug capacity, enabling effective hosting of therapeutics despite the integration complexity of its components, offering a promising solution for precise targeted drug delivery.

Different groups worldwide have been working with the GlobalFoundries 22nm platform (22FDX) with the hopes of industrializing the fabrication of Si spin qubits. To guide this effort, we have performed a systematic study of six of the foundry's processes of reference (POR). Using effective mobility as a figure of merit, we study the impact of gate stack, channel type and back bias as a function of temperature. This screening process selected qubit devices that allowed us to couple quantum dots along both the length and width of the Si channel. We present stability diagrams with clear and regular honeycomb patterns, where spurious elements such as dopants are not observed. By combining these results with room and low temperature simulations, we provide insights into potential technology optimizations and show both the utility of qubit pre-screening protocols as well as the advantages of leveraging forward body bias within an FDSOI (Fully Depleted Silicon-On-Insulator) qubit platform., Comment: 7 pages, 12 figures + 2 tables

We used a supervised machine learning algorithm (probabilistic random forest) to classify ~130 million sources in the VISTA Survey of the Magellanic Clouds (VMC). We used multi-wavelength photometry from optical to far-infrared as features to be trained on, and spectra of Active Galactic Nuclei (AGN), galaxies and a range of stellar classes including from new observations with the Southern African Large Telescope (SALT) and SAAO 1.9m telescope. We also retain a label for sources that remain unknown. This yielded average classifier accuracies of ~79% (SMC) and ~87% (LMC). Restricting to the 56,696,719 sources with class probabilities (P$_{class}$) > 80% yields accuracies of ~90% (SMC) and ~98% (LMC). After removing sources classed as 'Unknown', we classify a total of 707,939 (SMC) and 397,899 (LMC) sources, including > 77,600 extragalactic sources behind the Magellanic Clouds. The extragalactic sources are distributed evenly across the field, whereas the Magellanic sources concentrate at the centres of the Clouds, and both concentrate in optical/IR colour-colour/magnitude diagrams as expected. We also test these classifications using independent datasets, finding that, as expected, the majority of X-ray sources are classified as AGN (554/883) and the majority of radio sources are classed as AGN (1756/2694) or galaxies (659/2694), where the relative AGN-galaxy proportions vary substantially with radio flux density. We have found: > 49,500 hitherto unknown AGN candidates, likely including more AGN dust dominated sources which are in a critical phase of their evolution; > 26,500 new galaxy candidates and > 2800 new Young Stellar Object (YSO) candidates., Comment: MNRAS accepted

Recent advances in Large Language Models (LLMs) have demonstrated remarkable capabilities across various tasks with commercial models leading the way. While open models usually operate at a smaller scale, they maintain competitiveness through specialization and fine-tuning. However, a significant challenge persists: open models often underperform in low-resource languages due to limited representation in the training corpus. In this paper, we present LLMic, a bilingual foundation language model designed specifically for the Romanian Language. We document the complete process of pretraining a foundation model for a low-resource language, including corpus construction, architecture selection, and hyper-parameter optimization. Our evaluation demonstrates that LLMic can be specialized for tasks in the target language, achieving results comparable to other much larger open models. We show that fine-tuning LLMic for language translation after the initial pretraining phase outperforms existing solutions in English-to-Romanian translation tasks. This opens the path for efficient large-scale processing for the Romanian language community, using the much smaller LLMic model

Collectively-encoded qubits, involving ensembles of atomic or solid-state emitters, present many practical advantages for quantum technologies. However, they suffer from uncontrolled inhomogeneous dephasing which couples them to a quasi-continuum of dark states. In most cases, this process cannot be encompassed in a standard master equation with time-independent coefficients, making its description either tedious or inaccurate. We show that it can be understood as a displacement in time-frequency phase space and accurately included in resource-efficient numerical simulations of the qubit's dynamics. This description unveils a regime where the qubit becomes protected from dephasing through a combination of strong driving and non-Markovianity. We experimentally investigate this regime using a Rydberg superatom and extend its coherent dynamics beyond the inhomogeneous-dephasing characteristic time by an order of magnitude.

We consider Banach spaces $X$ that can be linearly lifted into their Lipschitz-free spaces $F(X)$ and, for a group $G$ acting on $X$ by linear isometries, we study the possible existence of $G$-equivariant linear liftings. In particular, we prove that such lifting exists when $G$ is compact in the strong operator topology, or an increasing union of such groups and $F(X)$ is complemented in its bidual by an equivariant projection.

We propose a method to deterministically entangle qubits or ensembles of qubits interacting with a shared bosonic mode in the ultrastrong coupling regime. We show that the resulting gate is a product of two unitaries: one unitary acts only on the quantum state of the qubits and entangles them, while the other acts only on the quantum state of the boson, producing a phase shift. We find that the gate time is inversely proportional to the qubit-boson interaction strength, and by tuning the qubit-boson interaction strength, one can prepare a maximally entangled state or a squeezed state. Applying the quantum gate to multiple qubit ensembles, we show that the quantum gate prepares a Schr\"odinger cat state. We also examine imperfections such as including free evolution of the qubits, and show that this produces an effective mixing. Our proposal is feasible for ultrastrong coupling experiments., Comment: 13 pages, 7 figures

We present VIRAC2, a catalogue of positions, proper motions, parallaxes and $Z$, $Y$, $J$, $H$, and $K_s$ near-infrared photometric time series of 545 346 537 unique stars. The catalogue is based on a point spread function fitting reduction of nearly a decade of VISTA VVV and VVVX images, which cover $560~{\rm deg}^2$ of the Southern Galactic plane and bulge. The catalogue is complete at the $>90$ per cent level for $11

Accurate load forecasting is crucial for predictive control in many energy domain applications, with significant economic and ecological implications. To address these implications, this study provides an extensive benchmark of state-of-the-art deep learning models for short-term load forecasting in energy communities. Namely, LSTM, xLSTM, and Transformers are compared with benchmarks such as KNNs, synthetic load models, and persistence forecasting models. This comparison considers different scales of aggregation (e.g., number of household loads) and varying training data availability (e.g., training data time spans). Further, the impact of transfer learning from synthetic (standard) load profiles and the deep learning model size (i.e., parameter count) is investigated in terms of forecasting error. Implementations are publicly available and other researchers are encouraged to benchmark models using this framework. Additionally, a comprehensive case study, comprising an energy community of 50 households and a battery storage demonstrates the beneficial financial implications of accurate predictions. Key findings of this research include: (1) Simple persistence benchmarks outperform deep learning models for short-term load forecasting when the available training data is limited to six months or less; (2) Pretraining with publicly available synthetic load profiles improves the normalized Mean Absolute Error (nMAE) by an average of 1.28%pt during the first nine months of training data; (3) Increased aggregation significantly enhances the performance of deep learning models relative to persistence benchmarks; (4) Improved load forecasting, with an nMAE reduction of 1.1%pt, translates to an economic benefit of approximately 600EUR per year in an energy community comprising 50 households., Comment: This preprint was submitted to the Elsevier journal Energy and AI on December 18, 2024

The study of regular black holes and black hole mimickers as alternatives to standard black holes has recently gained significant attention, driven both by the need to extend general relativity to describe black hole interiors, and by recent advances in observational technologies. Despite considerable progress in this field, significant challenges remain in identifying and characterizing physically well-motivated classes of regular black holes and black hole mimickers. This report provides an overview of these challenges, and outlines some of the promising research directions -- as discussed during a week-long focus programme held at the Institute for Fundamental Physics of the Universe (IFPU) in Trieste from November 11th to 15th, 2024., Comment: 39 pages, 3 figures

Speculative sampling is a popular technique for accelerating inference in Large Language Models by generating candidate tokens using a fast draft model and accepting or rejecting them based on the target model's distribution. While speculative sampling was previously limited to discrete sequences, we extend it to diffusion models, which generate samples via continuous, vector-valued Markov chains. In this context, the target model is a high-quality but computationally expensive diffusion model. We propose various drafting strategies, including a simple and effective approach that does not require training a draft model and is applicable out of the box to any diffusion model. Our experiments demonstrate significant generation speedup on various diffusion models, halving the number of function evaluations, while generating exact samples from the target model.

T Tauri stars and their discs are crucial for understanding stellar evolution and the formation of planets in low-mass systems. These stars exhibit significant variability, notably emitting intense X-ray flares due to magnetic reconnection events. During these events, magnetic energy is converted into kinetic energy of particles. Some of these particles then heat the plasma of the underlying chromosphere, emitting the observed X-rays. A portion of particles are thought to escape the chromosphere to interact with the surrounding circumstellar environment. The question is what impact do particles produced by magnetic reconnection events have on the discs of young stars. The complex characteristics of protoplanetary discs around T Tauri stars require an interdisciplinary strategy to enhance our understanding of these objects. This thesis contribute to establish a framework combining observational methodologies, chemical and dynamic models of protoplanetary discs, and the mechanics of acceleration and transport of energetic particles. We indroduce the role of ionisation in disc dynamics, including its sources. Not only from standard sources like stellar radiation and galactic cosmic rays but also from non-thermal ionisation due to magnetic reconnection events. We then examine energetic particles accelerated during magnetic reconnection events in T Tauri flares as an alternative source of ionisation, requiring the construction of a theoretical model based on solar flares. Then, we present a study revealing that particles from magnetic reconnection events could significantly contribute to the ionisation of the inner disc. Finally, we present a complementary study accounting for temporal effects, showing that considering these particles could increase the ionisation rate as well as viscosity, accretion rate, volumetric heating rate, and chemical complexity of inner protoplanetary discs., Comment: Doctoral thesis

Physarum polycephalum, a biological transport network, has emerged as a model for the study of self-organisation and coordination at large spatial scales. This self-organisation is largely due to cytoplasmic fluid flows driven by propagating contractile waves that are delocalised over the entire organism. Furthermore, complex modulation of these contractile waves on long time scales has also been observed, but is scarcely studied. In this study, we comprehensively characterise this long-time scale modulation by confining the organism within ring geometries.Our findings reveal a clear correlation between contractile wave direction, amplitude modulation, and the moving mean vein diameter. We observe simultaneous spatio-temporal travelling and standing wave patterns that bear a resemblance to the ring oscillating fundamental modes. Furthermore, we ascertain that the time periods of these modes scale with the system size and correspond to multiples of the statistically observed intrinsic modulation time scale. This suggests the presence of a common mechanism. These findings serve to reinforce the hypothesis that the flows patterns modulation is the result of the transport of a chemical agent responsible for modifying the local rigidity of the membrane, which would act as a signaling mechanism for the organism.

We consider chemical graphs that are defined as connected graphs of maximum degree at most 3. We characterize the extremal graphs, meaning those that maximize or minimize 33 degree-based topological indices. This study shows that five graph families are sufficient to characterize the extremal graphs of 29 of these 33 indices. In other words, the extremal properties of this set of degree-based topological indices vary very little., Comment: 25 pages

Effective image post-processing algorithms are vital for the successful direct imaging of exoplanets. Standard PSF subtraction methods use techniques based on a low-rank approximation to separate the rotating planet signal from the quasi-static speckles, and rely on signal-to-noise ratio maps to detect the planet. These steps do not interact or feed each other, leading to potential limitations in the accuracy and efficiency of exoplanet detection. We aim to develop a novel approach that iteratively finds the flux of the planet and the low-rank approximation of quasi-static signals, in an attempt to improve upon current PSF subtraction techniques. In this study, we extend the standard L2 norm minimization paradigm to an L1 norm minimization framework to better account for noise statistics in the high contrast images. Then, we propose a new method, referred to as Alternating Minimization Algorithm with Trajectory, that makes a more advanced use of estimating the low-rank approximation of the speckle field and the planet flux by alternating between them and utilizing both L1 and L2 norms. For the L1 norm minimization, we propose using L1 norm low-rank approximation, a low-rank approximation computed using an exact block-cyclic coordinate descent method, while we use randomized singular value decomposition for the L2 norm minimization. Additionally, we enhance the visibility of the planet signal using a likelihood ratio as a postprocessing step. Numerical experiments performed on a VLT/SPHERE-IRDIS dataset show the potential of AMAT to improve upon the existing approaches in terms of higher S/N, sensitivity limits, and ROC curves. Moreover, for a systematic comparison, we used datasets from the exoplanet data challenge to compare our algorithm to other algorithms in the challenge, and AMAT with likelihood ratio map performs better than most algorithms tested on the exoplanet data challenge., Comment: Submitted to A&A, 20 pages, 20 figures