Search

This paper explores the optimization of teacher education programs to meet the evolving demands of 21st-century classrooms. It examines key strategies to enhance teacher preparation, such as the integration of technology, fostering critical thinking, promoting inclusivity, and preparing educators to address global challenges. The paper also explores theoretical frameworks such as TPACK (Technological Pedagogical Content Knowledge) and constructivism to guide the integration of innovative pedagogies. Drawing on empirical studies and current literature, the paper suggests practical approaches for reimagining teacher education to prepare teachers for dynamic, diverse, and technology-driven learning environments.

The exponential growth of online learning has catalyzed significant pedagogical innovations and transformed the educational landscape. This paper explores the emerging trends in online learning, including the shift towards blended learning, the rise of personalized learning, and the integration of technology-enhanced pedagogical practices. The study also examines global perspectives on online education, the role of new technologies in facilitating innovative learning models, and the pedagogical shifts that have accompanied the rise of online learning. The paper concludes by offering insights into the future directions of online education and the ongoing challenges faced by educators and learners in this digital age.

This paper explores innovative approaches to curriculum design and teacher education that aims to address the evolving needs of 21st-century learners. With the increasing integration of technology, the need for teacher professional development and the importance of inclusive education, this paper examines key trends in curriculum development and teacher education. It also emphasizes the need for pedagogical practices that foster critical thinking, creativity, and inclusivity, while preparing students for a globalized world.

The rapidly evolving global economy demands a shift in educational priorities. This paper emphasizes the need for African curricula to integrate future skills such as digital literacy, critical thinking, collaboration, and adaptability. By redesigning curriculum frameworks, African nations can empower students to thrive in competitive international job markets. However, systemic challenges such as teacher training, infrastructure deficits, and cultural context must be addressed to ensure the effective implementation of these changes.

Artificial Intelligence (AI) is revolutionizing industries, yet its integration into education remains underutilized. This paper advocates for embedding AI tools in curriculum design to personalize learning experiences, address diverse student needs, and equip learners with future-ready skills. By leveraging AI's capabilities, educators can create adaptive, inclusive, and dynamic curricula, preparing students for the demands of a technology-driven world. However, this integration requires careful consideration of ethical challenges, infrastructure demands, and capacity building for educators.

This paper advocates for the integration of Social-Emotional Learning (SEL) into educational curricula to foster holistic student development. SEL equips students with emotional intelligence, resilience, and interpersonal skills, which are crucial for personal and professional success. As the global focus on mental health and emotional well-being grows, embedding SEL in educational systems offers a proactive approach to addressing these priorities while creating well-rounded individuals capable of navigating complex social and emotional landscapes.

This dissertation investigates the interplay of academic preparedness, imposter syndrome, and resilience among low-income male African American students, focusing on their readiness for college and subsequent success. Using a phenomenological approach, the study captures the lived experiences of these students and reveals systemic challenges and psychological barriers. Findings indicate that disparities in academic preparedness, including unequal access to quality education and experienced teachers, significantly impact college readiness. Imposter syndrome is shown to diminish academic confidence and performance, while resilience proves essential in overcoming these challenges. The study also evaluates the effectiveness of support programs like GEAR UP and TRIO, demonstrating their critical role in promoting college preparedness. The research concludes that tailored interventions addressing the unique needs of male African American students are necessary for fostering academic success and equity. Implications for policy include advocating for modifications to ensure equal access to rigorous curricula, skilled teachers, and adequate guidance counseling. This study contributes to the broader discourse on educational equity and resilience in overcoming academic barriers.

This paper presents causal evidence on the impact of parenting practices on early child development. We exploit exogenous changes in nurturing care induced by a parent training intervention to estimate the impact of nurturing parenting practices on child outcomes. We find a large and significant impact measured at age two; in contrast, at age four nurturing care has only a modest, and imprecisely estimated, impact on child outcomes. This is despite the fact that the intervention induced substantial changes in parenting practices at both ages. The differential relationship between child development and nurturing care at ages two and four explains the fade-out in treatment effects for the intervention as a whole: although parents continued to respond, their response no longer had the intended effect on child outcomes. [Additional funding from the Royal Society of Edinburgh.]

Test-time scaling is a promising new approach to language modeling that uses extra test-time compute to improve performance. Recently, OpenAI's o1 model showed this capability but did not publicly share its methodology, leading to many replication efforts. We seek the simplest approach to achieve test-time scaling and strong reasoning performance. First, we curate a small dataset s1K of 1,000 questions paired with reasoning traces relying on three criteria we validate through ablations: difficulty, diversity, and quality. Second, we develop budget forcing to control test-time compute by forcefully terminating the model's thinking process or lengthening it by appending "Wait" multiple times to the model's generation when it tries to end. This can lead the model to double-check its answer, often fixing incorrect reasoning steps. After supervised finetuning the Qwen2.5-32B-Instruct language model on s1K and equipping it with budget forcing, our model s1 exceeds o1-preview on competition math questions by up to 27% (MATH and AIME24). Further, scaling s1 with budget forcing allows extrapolating beyond its performance without test-time intervention: from 50% to 57% on AIME24. Our model, data, and code are open-source at https://github.com/simplescaling/s1., Comment: 46 pages (9 main), 10 figures, 14 tables

In this article we use the latest cosmological observations, including SNe, BAO, CC and RSD, to reconstruct the cosmological evolution via the Gaussian process. At the background level, we find consistency with the quintom dynamics for different data combinations and divide the characteristics of dark energy into three different categories, which are negative-energy dark energy, late-dominated dark energy and oscillating dark energy, respectively. Considering the effect of modified gravity on the growth of matter perturbations, the reconstruction results at the perturbative level show that we only need minor corrections to general relativity. Furthermore, we provide theoretical interpretation for the three different types of dynamical dark-energy behavior, in the framework of modified gravity, scalar fields, and dark-energy equation-of-state parametrizations. Finally, we show that all of these models can be unified in the framework of effective field theory., Comment: 19 pages, 6 figures

The insatiable appetite of Artificial Intelligence (AI) workloads for computing power is pushing the industry to develop faster and more efficient accelerators. The rigidity of custom hardware, however, conflicts with the need for scalable and versatile architectures capable of catering to the needs of the evolving and heterogeneous pool of Machine Learning (ML) models in the literature. In this context, multi-chiplet architectures assembling multiple (perhaps heterogeneous) accelerators are an appealing option that is unfortunately hindered by the still rigid and inefficient chip-to-chip interconnects. In this paper, we explore the potential of wireless technology as a complement to existing wired interconnects in this multi-chiplet approach. Using an evaluation framework from the state-of-the-art, we show that wireless interconnects can lead to speedups of 10% on average and 20% maximum. We also highlight the importance of load balancing between the wired and wireless interconnects, which will be further explored in future work., Comment: Accepted in AccML @ HiPEAC 2025

Structural anisotropy in layered two-dimensional materials can lead to highly anisotropic optical absorption which, in turn, can profoundly a^ect their phonon modes. These e^ects include lattice orientation-dependent and excitation energy-dependent mode intensities that can enable new phononic and optoelectronic applications. Here, we report anomalous Raman spectra in single-crystalline AgCrP$_2$Se$_6$, a layered antiferromagnetic material. Density functional theory calculations and experimental measurements reveal several unique features in the Raman spectra of bulk and exfoliated AgCrP$_2$Se$_6$ crystals including three helical vibrational modes. These modes exhibit large Raman optical activities (circular intensity di^erences) in bulk AgCrP$_2$Se$_6$, which progressively decrease with thickness. We also observe strong excitation energy dependent peak intensities as well as a decrease in anti-Stokes peak intensities at room temperature with increasing excitation energy, resulting in an apparent cooling by up to 220 K. All of these anomalies in bulk and exfoliated flakes are attributed to the unique ABC layer stacking structure of AgCrP$_2$Se$_6$ and to the smaller unit cell volume that causes hybridization between the Se and Ag/Cr electron densities, resulting in charge transfer and strongly a^ecting the electron-phonon coupling. This work thus positions AgCrP$_2$Se$_6$ as an exciting new 2D material for optical and phononic applications., Comment: 22 pages, 6 figures

While shared autonomy offers significant potential for assistive robotics, key questions remain about how to effectively map 2D control inputs to 6D robot motions. An intuitive framework should allow users to input commands effortlessly, with the robot responding as expected, without users needing to anticipate the impact of their inputs. In this article, we propose a dynamic input mapping framework that links joystick movements to motions on control frames defined along a trajectory encoded with canal surfaces. We evaluate our method in a user study with 20 participants, demonstrating that our input mapping framework reduces the workload and improves usability compared to a baseline mapping with similar motion encoding. To prepare for deployment in assistive scenarios, we built on the development from the accessible gaming community to select an accessible control interface. We then tested the system in an exploratory study, where three wheelchair users controlled the robot for both daily living activities and a creative painting task, demonstrating its feasibility for users closer to our target population.

Some white dwarfs undergo significant changes in atmospheric composition owing to the diffusion and mixing of residual hydrogen in a helium-rich envelope. Of particular interest are a few objects exhibiting hydrogen and helium line variations modulated by rotation, revealing surface composition inhomogeneities. Recently, the hot ultramassive white dwarf ZTF J203349.80+322901.1 emerged as the most extreme such specimen, with hydrogen and helium lines successively appearing and vanishing in anti-phase, suggesting a peculiar double-faced configuration. However, standard atmosphere models fail to reproduce the observed spectrum at all rotation phases, hampering further interpretation. Here, we perform a new analysis of ZTF J203349.80+322901.1 using stratified atmosphere models, where hydrogen floats above helium, and obtain excellent fits to the phase-resolved spectra. Our results imply that an extremely thin hydrogen layer covers the entire surface but varies from optically thick to optically thin across the surface, thus producing the observed spectral variations. We present new envelope models indicating that the hydrogen layer arises from a delicate interplay between diffusion and convection. We discuss possible explanations for the surface layer asymmetry, including an asymmetric magnetic field and a non-uniform internal hydrogen distribution. Finally, we highlight implications for expanding and understanding the emerging class of inhomogeneous white dwarfs., Comment: 7 pages, 4 figures, accepted for publication in MNRAS Letters

We present how Wald-Gauss-Bonnet topological dark energy can alleviate $H_{0}$ and $\sigma 8$ tensions simultaneously. This modified cosmological scenario is obtained by applying the gravity-thermodynamics conjecture to the Universe apparent horizon, employing the Wald-Gauss-Bonnet entropy instead of the standard Bekenstein Hawking one. Therefore, the modified Friedmann equations and the effective dark energy depend on the Gauss-Bonnet coupling constant and on measured astrophysical parameters such as the black-hole formation rate and the black-hole progenitors average mass. Wald-Gauss-Bonnet topological dark energy exhibits phantom behavior at intermediate redshifts for positive coupling constants, which is known to be a sufficient mechanism that can lead to the $H_{0}$ tension alleviation. Indeed, we show that for particular regions of the parameter space the scenario at hand does lead to higher $H_{0}$ values, and we verify this result performing a confrontation with observational data from Supernovae Type Ia (SNIa) and Cosmic Chronometers (CC) measurements. Additionally, analyzing the growth of structures and the evolution of the matter overdensity we show that the redshift behavior of the $f\sigma_{8}$ observable obtains smaller values than the ones arising in $\Lambda$CDM cosmology, and thus the scenario at hand can alleviate the $\sigma_{8}$ tension too. This is an advantage of Wald-Gauss-Bonnet cosmology, since the simultaneous alleviation of both tensions cannot be easily achieved within other entropic scenarios.

The global adoption of Large Language Models (LLMs) in healthcare shows promise to enhance clinical workflows and improve patient outcomes. However, Automatic Speech Recognition (ASR) errors in critical medical terms remain a significant challenge. These errors can compromise patient care and safety if not detected. This study investigates the prevalence and impact of ASR errors in medical transcription in Nigeria, the United Kingdom, and the United States. By evaluating raw and LLM-corrected transcriptions of accented English in these regions, we assess the potential and limitations of LLMs to address challenges related to accents and medical terminology in ASR. Our findings highlight significant disparities in ASR accuracy across regions and identify specific conditions under which LLM corrections are most effective., Comment: 15 pages, 8 figures

Multimode lasers have a very complex dynamics, as expected when oscillators are nonlinearly coupled. Order emerges when the modes lock together; in this case the coherent superposition of the modes results into a periodic train of pulses or a nearly constant power output with a linearly chirped frequency, for instance. The first is promoted by a saturable absorber, or an equivalent physical mechanism, while the latter is connected to more subtle conditions, such as the fast dynamics of the gain. Here we consider the case of a multimode semiconductor laser with gain provided by quantum wells but without any saturable absorber. The cavity is designed to have a photonic bandgap and very low dispersion. We show, first in theory, that modes can lock together and generate a variety of waveforms which are not trains of pulses nor chirped continuous power waves. Mode locking is observed in experiments on a III-V/Silicon hybrid laser with the cavity made of a suitably tapered grating. Moreover, we find that the mode-locking beatnote is strongly dependent on the injected current: we reach more than 1 GHz modulation amplitude of the beatnote at a modulation frequency of 50 kHz. The behaviour of the laser is critically determined by the dispersion, which can be controlled by the photonic crystal structure. By scaling up the number of interacting modes, this laser source may offer an effective and extremely flexible way of generating waveforms \`a la carte., Comment: 8 pages, 5 figures

Zinc phosphorus trisulfide (ZnPS$_3$), a promising material for photocatalysis and energy storage, is shown in this study to exhibit remarkable stability under extreme conditions. We explore its optical and structural properties under high pressure and cryogenic temperatures using photoluminescence (PL) spectroscopy, Raman scattering, and density functional theory (DFT). Our results identify a pressure-induced phase transition starting at 6.75 GPa and stabilizing by 12.5 GPa, after which ZnPS$_3$ demonstrates robust stability across a broad pressure range of 15 to 100 GPa. DFT calculations predict a semiconductor-to-semimetal transition at 100 GPa, while PL measurements reveal defect-assisted emissions that quench under pressure due to enhanced non-radiative recombination. At cryogenic temperatures, PL quenching intensifies as non-radiative processes dominate, driven by a rising Gr\"uneisen parameter and reduced phonon population. Cryogenic X-ray diffraction (XRD) also reveals a high mean thermal expansion coefficient (TEC) of (4.369 $\pm$ 0.393) $\times$ 10$^{-5}$ K$^{-1}$, among the highest reported for 2D materials. This unique combination of tunable electronic properties under low pressure and high thermal sensitivity makes ZnPS$_3$ a strong candidate for sensing applications in extreme environments.

Domain generalization aims to find ways for deep learning models to maintain their performance despite significant domain shifts between training and inference datasets. This is particularly important for models that need to be robust or are costly to train. LiDAR perception in autonomous driving is impacted by both of these concerns, leading to the emergence of various approaches. This work addresses the challenge by proposing a geometry-based approach, leveraging the sequential structure of LiDAR sensors, which sets it apart from the learning-based methods commonly found in the literature. The proposed method, called 3DLabelProp, is applied on the task of LiDAR Semantic Segmentation (LSS). Through extensive experimentation on seven datasets, it is demonstrated to be a state-of-the-art approach, outperforming both naive and other domain generalization methods.

We present an adaptation of the so-called structural method \cite{CMM23} for Hamiltonian systems, and redesign the method for this specific context, which involves two coupled differential systems. Structural schemes decompose the problem into two sets of equations: the physical equations, which describe the local dynamics of the system, and the structural equations, which only involve the discretization on a very compact stencil. They have desirable properties, such as unconditional stability or high-order accuracy. We first give a general description of the scheme for the scalar case (which corresponds to e.g. spring-mass interactions or pendulum motion), before extending the technique to the vector case (treating e.g. the $n$-body system). The scheme is also written in the case of a non-separable system (e.g. a charged particle in an electromagnetic field). We give numerical evidence of the method's efficiency, its capacity to preserve invariant quantities such as the total energy, and draw comparisons with the traditional symplectic methods., Comment: 35 pages, 7 figures

Chronic wounds affect 8.5 million Americans, particularly the elderly and patients with diabetes. These wounds can take up to nine months to heal, making regular care essential to ensure healing and prevent severe outcomes like limb amputations. Many patients receive care at home from visiting nurses with varying levels of wound expertise, leading to inconsistent care. Problematic, non-healing wounds should be referred to wound specialists, but referral decisions in non-clinical settings are often erroneous, delayed, or unnecessary. This paper introduces the Deep Multimodal Wound Assessment Tool (DM-WAT), a machine learning framework designed to assist visiting nurses in deciding whether to refer chronic wound patients. DM-WAT analyzes smartphone-captured wound images and clinical notes from Electronic Health Records (EHRs). It uses DeiT-Base-Distilled, a Vision Transformer (ViT), to extract visual features from images and DeBERTa-base to extract text features from clinical notes. DM-WAT combines visual and text features using an intermediate fusion approach. To address challenges posed by a small and imbalanced dataset, it integrates image and text augmentation with transfer learning to achieve high performance. In evaluations, DM-WAT achieved 77% with std 3% accuracy and a 70% with std 2% F1 score, outperforming prior approaches. Score-CAM and Captum interpretation algorithms provide insights into specific parts of image and text inputs that influence recommendations, enhancing interpretability and trust., Comment: arXiv admin note: text overlap with arXiv:2208.05051 by other authors

The rapid proliferation of AI models, coupled with growing demand for edge deployment, necessitates the development of AI hardware that is both high-performance and energy-efficient. In this paper, we propose a novel analog accelerator architecture designed for AI/ML training workloads using stochastic gradient descent with L2 regularization (SGDr). The architecture leverages log-domain circuits in subthreshold MOS and incorporates volatile memory. We establish a mathematical framework for solving SGDr in the continuous time domain and detail the mapping of SGDr learning equations to log-domain circuits. By operating in the analog domain and utilizing weak inversion, the proposed design achieves significant reductions in transistor area and power consumption compared to digital implementations. Experimental results demonstrate that the architecture closely approximates ideal behavior, with a mean square error below 0.87% and precision as low as 8 bits. Furthermore, the architecture supports a wide range of hyperparameters. This work paves the way for energy-efficient analog AI hardware with on-chip training capabilities.

We present a method based on optimal transport to remove arbitrage opportunities within a finite set of option prices. The method is notably intended for regulatory stress-tests, which impose to apply important local distortions to implied volatility surfaces. The resulting stressed option prices are naturally associated to a family of signed marginal measures: we formulate the process of removing arbitrage as a projection onto the subset of martingale measures with respect to a Wasserstein metric in the space of signed measures. We show how this projection problem can be recast as an optimal transport problem; in view of the numerical solution, we apply an entropic regularization technique. For the regularized problem, we derive a strong duality formula, show convergence results as the regularization parameter approaches zero, and formulate a multi-constrained Sinkhorn algorithm, where each iteration involves, at worse, finding the root of an explicit scalar function. The convergence of this algorithm is also established. We compare our method with the existing approach by [Cohen, Reisinger and Wang, Appl.\ Math.\ Fin.\ 2020] across various scenarios and test cases.

In this paper, we prove that sufficiently regular solutions of any quasilinear PDE can be approximated by solutions of systems of N distinguishable particles, to within 1/ ln(N ). This intruiguing result is related to recent developments in graph limit theory.

Axion-like particles (ALP) are promising candidates to comprise all the dark matter in the universe. We investigate the ALP couplings to photons and electrons via astrophysical measurements through the search for very-high-energy gamma rays arising from high-energy cosmic-ray scattering off ALP populating the halo of the Milky Way. We show that gamma-ray signals from ALP couplings to photons and electrons via inverse Primakoff and Compton processes respectively, can be probed by very-high-energy ($\gtrsim$100 GeV) gamma-ray ground-based observatories, providing an alternative and complementary avenue to probe ALP couplings in the eV mass range. Sensitivities of current and near-future ground-based gamma-ray observatories improves upon one order of magnitude the current constraints from gamma-ray satellite experiments for the ALP-photon couplings in the region of masses below 10$^{-9}$ GeV. Their sensitivities reached on the ALP-electron couplings allow probing masses below 10$^{-8}$ GeV, which are lower than the masses probed in gamma-ray satellite experiments., Comment: 11 pages, 5 figures. Accepted for publication in PRD

The Internet of Things (IoT) is a network of digital devices like sensors, processors, embedded and communication devices that can connect to and exchange data with other devices and systems over the internet. IoT devices have limitations on power, memory, and computational resources. Researchers have developed the IPv6 Over Low-power Wireless Personal Area Network (6LoWPAN) protocols to provide wireless connectivity among these devices while overcoming the constraints on resources. 6LoWPAN has been approved subsequently by the Internet Engineering Task Force (IETF). The IETF Routing Over Low-power and Lossy Networks (ROLL) standardized the Routing Protocol for LLNs known as RPL (IETF RFC 6550), which is part of the 6LoWPAN stack. However, IoT devices are vulnerable to various attacks on RPL-based routing. This survey provides an in depth study of existing RPL-based attacks and defense published from year 2011 to 2024 from highly reputed journals and conferences. By thematic analysis of existing routing attacks on RPL, we developed a novel attack taxonomy which focuses on the nature of routing attacks and classifies them into 12 major categories. Subsequently, the impact of each attack on the network is analyzed and discussed real life scenarios of these attacks. Another contribution of this survey proposed a novel taxonomy for classification of defense mechanisms into 8 major categories against routing attacks based on type of defense strategy. The detailed analysis of each defense mechanism with real life applicability is explained. Furthermore, evaluation tools such as testbeds and simulators for RPL-based attack and defense are discussed and critically analyzed in terms of real world applicability. Finally, open research challenges are presented on the basis of research gaps of existing literature along with research directions for practitioners and researchers.

EllipSect is a user-friendly Python analysis tool that generates publication quality figures from the output of GALFIT 3, a popular galaxy surface brightness 2D-fitting algorithm. EllipSect allows the direct analysis of GALFIT 3 preliminary modeling, to guide optimal parameter selection. Furthermore, EllipSect generates non-parametric measurements of some useful parameters that are not provided by GALFIT 3, such as the effective radius resulting from multi-component fits, cusp radius, and the Petrosian radius. This paper provides examples and a quick guide for EllipSect., Comment: 20 pages, 7 figures, submitted to RevMexAA

We present Mantis Shrimp, a multi-survey deep learning model for photometric redshift estimation that fuses ultra-violet (GALEX), optical (PanSTARRS), and infrared (UnWISE) imagery. Machine learning is now an established approach for photometric redshift estimation, with generally acknowledged higher performance in areas with a high density of spectroscopically identified galaxies over template-based methods. Multiple works have shown that image-based convolutional neural networks can outperform tabular-based color/magnitude models. In comparison to tabular models, image models have additional design complexities: it is largely unknown how to fuse inputs from different instruments which have different resolutions or noise properties. The Mantis Shrimp model estimates the conditional density estimate of redshift using cutout images. The density estimates are well calibrated and the point estimates perform well in the distribution of available spectroscopically confirmed galaxies with (bias = 1e-2), scatter (NMAD = 2.44e-2) and catastrophic outlier rate ($\eta$=17.53$\%$). We find that early fusion approaches (e.g., resampling and stacking images from different instruments) match the performance of late fusion approaches (e.g., concatenating latent space representations), so that the design choice ultimately is left to the user. Finally, we study how the models learn to use information across bands, finding evidence that our models successfully incorporates information from all surveys. The applicability of our model to the analysis of large populations of galaxies is limited by the speed of downloading cutouts from external servers; however, our model could be useful in smaller studies such as generating priors over redshift for stellar population synthesis.

Artificial neural networks and their applications in deep learning have recently made an incursion into the field of control. Deep learning techniques in control are often related to optimal control, which relies on Pontryagin maximum principle or the Hamilton-Jacobi-Bellman equation. They imply control schemes that are tedious to implement. We show here that the new HEOL setting, resulting from the fusion of the two established approaches, namely differential flatness and model-free control, provides a solution to control problems that is more sober in terms of computational resources. This communication is devoted to the synchronization of the popular Kuramoto's coupled oscillators, which was already considered via artificial neural networks (B\"ottcher et al., Nature Communications 2022), where, contrarily to this communication, only the single control variable is examined. One establishes the flatness of Kuramoto's coupled oscillator model with multiplicative control and develops the resulting HEOL control. Unlike many exemples, this system reveals singularities that are avoided by a clever generation of phase angle trajectories. The results obtained, verified in simulation, show that it is not only possible to synchronize these oscillators in finite time, and even to follow angular frequency profiles, but also to exhibit robustness concerning model mismatches. To the best of our knowledge this has never been done before. Concluding remarks advocate a viewpoint, which might be traced back to Wiener's cybernetics: control theory belongs to AI., Comment: MATHMOD 2025 (11th Vienna International Conference on Mathematical Modelling, 19-21 February 2025, Vienna, Austria)

With enough X-ray flux to be detected in a 160s scan by SRG/eROSITA, the $z = 6.19$ quasar CFHQS J142952+544717 is, by far, the most luminous X-ray source known at $z > 6$. We present deep (245 ks) NuSTAR observations of this source; with $\sim180$ net counts in the combined observations, CFHQS J142952+544717 is the most distant object ever observed by the observatory. Fortuitously, this source was independently observed by Chandra $\sim110$ days earlier, enabling the identification of two nearby (30'' and 45'' away), fainter X-ray sources. We jointly fit both Chandra and NuSTAR observations--self-consistently including interloper sources--and find that, to greater than 90% confidence, the observed 3-7 keV flux varied by a factor of $\sim2.6$ during that period, corresponding to approximately two weeks in the quasar rest-frame. This brightening is one the most extreme instances of statistically significant X-ray variability seen in the Epoch of Reionization. We discuss possible scenarios that could produce such rapid change, including X-ray emission from jets too faint at radio frequencies to be observed., Comment: 16 pages, 7 Figures, 3 Tables; accepted for publication in the Astrophysical Journal Letters (ApJL)

We report on the characteristics of a microwave photon counter device based on a superconducting transmon qubit. Its design is similar to [arXiv:2307.03614], with an additional bandwidth tuning circuit that allows optimizing the device efficiency and noise. Owing to this new feature and to improvements in device fabrication, a power sensitivity of $3 \cdot 10^{-23} \mathrm{W}/\sqrt{\mathrm{Hz}}$ is reached. We confirm the high performance of the device by measuring single spin microwave fluorescence., Comment: 12 pages, 10 figures

We consider (random) walks in a multidimensional orthant. Using the idea of universality in probability theory, one can associate a unique polyhedral domain to any given walk model. We use this connection to prove two sets of new results. First, we are interested in a group of transformations naturally associated with any small step model; as it turns out, this group is central to the classification of walk models. We show a strong connection between this group and the reflection group through the walls of the polyhedral domain. As a consequence, we can derive various conditions for the combinatorial group to be infinite. Secondly, we consider the asymptotics of the number of excursions, whose critical exponent is known to be computable in terms of the eigenvalue of the above polyhedral domain. We prove new results from spectral theory on the eigenvalues of polyhedral nodal domains. We believe that these results are interesting in their own right; they can also be used to find new exact asymptotic results for walk models corresponding to these nodal polyhedral domains., Comment: 30 pages

We define a positive state sum for webs "of type D". These webs are graphs which mimic morphisms in the category of finite-dimensional quantum so(2N)-modules. From the state sum, we derive an invariant of framed unoriented links. After giving explicit details about some intertwiners in the category of quantum so(2N)-modules, we relate our state-sum link invariant with Reshetikhin--Turaev's invariant associated with quantum so(2N)., Comment: 48 pages, many diagrams, comments welcome!

In this paper, we investigate on a bounded open set of $\mathbb{R}^N$ with smooth boundary, an eigenvalue problem involving the sum of nonlocal operators $(-\Delta)_p^{s_1}+ (-\Delta)_q^{s_2}$ with $s_1,s_2\in (0,1)$, $p,q\in (1,\infty)$ and subject to the corresponding homogeneous nonlocal $(p,q)$-Neumann boundary condition. A careful analysis of the considered problem leads us to a complete description of the set of eigenvalues as being the precise interval $\{0\}\cup(\lambda_{1}(s_2,q),\infty)$, where $\lambda_{1}(s_2,q)$ is the first nonzero eigenvalue of the homogeneous fractional $q$-Laplacian under nonlocal $q$-Neumann boundary condition. Furthermore, we establish that every eigenfunctions is globally bounded., Comment: 26 pages. arXiv admin note: text overlap with arXiv:2406.15825

Convex functions and their gradients play a critical role in mathematical imaging, from proximal optimization to Optimal Transport. The successes of deep learning has led many to use learning-based methods, where fixed functions or operators are replaced by learned neural networks. Regardless of their empirical superiority, establishing rigorous guarantees for these methods often requires to impose structural constraints on neural architectures, in particular convexity. The most popular way to do so is to use so-called Input Convex Neural Networks (ICNNs). In order to explore the expressivity of ICNNs, we provide necessary and sufficient conditions for a ReLU neural network to be convex. Such characterizations are based on product of weights and activations, and write nicely for any architecture in the path-lifting framework. As particular applications, we study our characterizations in depth for 1 and 2-hidden-layer neural networks: we show that every convex function implemented by a 1-hidden-layer ReLU network can be also expressed by an ICNN with the same architecture; however this property no longer holds with more layers. Finally, we provide a numerical procedure that allows an exact check of convexity for ReLU neural networks with a large number of affine regions.

In this study, we investigate the structural properties of a cyclic and linear polymer blend confined within interfaces, comparing the properties of the different polymer systems as their composition and length vary through molecular dynamics simulations. Our analysis reveals that the local composition orderings of cyclic and linear polymers near the surface exhibit distinct differences. However, at longer lengths and increasing the cyclic concentration, it was shown that heat capacity increases until it gets to a threshold whereby increasing the length or cyclic concentration results in decrease and subsequently similar values. Notably, our findings indicate the stress-related properties of polymer blends and the influence of molecular weight and chain structure on developing sustainable strategies for lessening the environmental impact of polymers, Comment: 23 PAGES, 10 FIGURES

It is well-established that shear layer flows are linearly unstable provided the viscosity is small enough, when the horizontal Fourier number lies in some interval, between the so called lower and upper marginally stable curves. In this article, we prove that monotonic shear layers undergo a bifurcation near the upper marginally stable curve. In the particular case of the exponential boundary layer, we give evidences that this bifurcation is a Hopf bifurcation, which corresponds to the onset of moving rolls near the boundary. The onset of such rolls is the first step of one of the possible ``routes to chaos'' for shear layer flows.

Clouds play a key role in Earth's radiation balance with complex effects that introduce large uncertainties into climate models. Real-time 3D cloud data is essential for improving climate predictions. This study leverages geostationary imagery from MSG/SEVIRI and radar reflectivity measurements of cloud profiles from CloudSat/CPR to reconstruct 3D cloud structures. We first apply self-supervised learning (SSL) methods-Masked Autoencoders (MAE) and geospatially-aware SatMAE on unlabelled MSG images, and then fine-tune our models on matched image-profile pairs. Our approach outperforms state-of-the-art methods like U-Nets, and our geospatial encoding further improves prediction results, demonstrating the potential of SSL for cloud reconstruction.

The reconfiguration problem for homomorphisms of digraphs to a reflexive digraph cycle, which amounts to deciding if a `reconfiguration graph' is connected, is known to by polynomially time solvable via a greedy algorithm based on certain topological requirements. Even in the case that the instance digraph is a cycle of length $m$, the algorithm, being greedy, takes time $\Omega(m^2)$. Encoding homomorphisms between two cycles as a relation on strings that represent the orientations of the cycles, we give a characterization of the components of the reconfiguration graph that can be computed in linear time and logarithmic space. In particular, this solves the reconfiguration problem for homomorphisms of cycle to cycles in log-space., Comment: 14 pages

In this article, we construct unique strong solutions to a class of stochastic Volterra differential equations driven by a singular drift vector field and a Wiener noise. Further, we examine the Sobolev differentiability of the strong solution with respect to its initial value.

We apply the gravity-thermodynamics approach in the case of Einstein-Gauss-Bonnet theory, and its corresponding Wald-Gauss-Bonnet entropy, which due to the Chern-Gauss-Bonnet theorem it is related to the Euler characteristic of the Universe topology. However, we consider the realistic scenario where we have the formation and merger of black holes that lead to topology changes, which induce entropy changes in the Universe horizon. We extract the modified Friedmann equations and we obtain an effective dark energy sector of topological origin. We estimate the black-hole formation and merger rates starting from the observed star formation rate per redshift, which is parametrized very efficiently by the Madau-Dickinson form, and finally we result to a dark-energy energy density that depends on the cosmic star formation rate density, on the fraction $f_{\text{BH}}$ of stars forming black holes, on the fraction of black holes $f_\text{merge}$ that eventually merge, on the fraction $ f_{\text{bin}}$ of massive stars that are in binaries, on the average mass of progenitor stars that will evolve to form black holes $ \langle m_{\text{prog}} \rangle $, as well as on the Gauss-Bonnet coupling constant. We investigate in detail the cosmological evolution, obtaining the usual thermal history. Concerning the dark-energy equation-of-state parameter, we show that at intermediate redshifts it exhibits phantom-like or quintessence-like behavior according to the sign of the Gauss-Bonnet coupling, while at early and late times it tends to the cosmological constant value. Finally, we study the effect of the other model parameters, showing that for the whole allowed observationally estimated ranges, the topological dark-energy equation-of-state parameter remains within its observational bounds., Comment: 13 pages, 11 figures, 1 Table

We Study versions of Cauchy formula in more general algebras than the complex case., Comment: 53 pages, in French language

Homogeneity and efficient oxygen transfer are crucial for aerobic cultures, which is popularly performed in Stirred Tank Bioreactors, through internal mechanical agitation of the impellers.Although there are a number of impeller designs for achieving this purpose, there are still concerns about the ability of the impellers to yield homogeneity and mitigate or eliminate stagnant zones.In this study, a novel impeller design, with auxiliary agitators in form of vents, was introduced and evaluated for small lab-scale bioreactors. For the evaluation, 3D models of a single and double impeller configurations, placed in two different bioreactors were developed. Computational fluid dynamics was employed to carry out the hydrodynamic simulation using k-epsilon standard model in the bioreactors.Computational variables such as the flow velocity, streamlines, pressure and wall shear stress on the shaft and impellers, eddy viscosity, turbulence eddy dissipation and turbulence kinetic energy were obtained and compared in both bioreactors to evaluate the performances at speeds of 50, 100, and 150 revolutions per minute.A comparison of the results with traditional segment-segment and segment-Rushton impellers shows that our double impeller configuration performs more desirably at speeds ranging from 100 to 150 RPM. Homogeneity was also achieved in both bioreactors, and there was significant reduction of stagnant zone less than 99 percentage in the double impeller configuration and significant mitigation in the single impeller agitation., Comment: 28 pages, 16 figures. No conference

Coupled spins form composite quantum systems which play an important role in many quantum technology applications, with an essential task often being the efficient generation of entanglement between two constituent qubits. The simplest such system is a pair of spins-$1/2$ coupled with Ising interaction, and in previous works various quantum control methods such as adiabatic processes, shortcuts to adiabaticity and optimal control have been employed to quickly generate there one of the maximally entangled Bell states. In this study, we use machine learning and optimization methods to produce maximally entangled states in minimum time, with the Rabi frequency and the detuning used as bounded control functions. We do not target a specific maximally entangled state, like the preceding studies, but rather find the controls which maximize the concurrence, leading thus automatically the system to the closest such state in shorter time. By increasing the bounds of the control functions we observe that the corresponding optimally selected maximally entangled state also changes and the necessary time to reach it is reduced. The present work demonstrates also that machine learning and optimization offer efficient and flexible techniques for the fast generation of entanglement in coupled spin systems, and we plan to extent it to systems involving more spins, for example spin chains.

Mathematical models of photosynthesis-irradiance relationships in phytoplankton are used to compute integrated water-column photosynthetic rates and predict primary production in ecosystem models. Models typically ignore an important phenomenon observed in most experiments: photosynthetic rate is an approximately constant maximum value over a range of irradiances until photoinhibition leads to decreasing photosynthetic rate. Here we develop a new model of photoinhibition that captures this plateau. We test six new models of photoinhibition and ten more photoinhibition models from the literature against a database of 1808 photosynthesis-irradiance curves exhibiting photoinhibition. In the best model, photoinhibition is phenomenologically described by multiplication by a saturating function of the reciprocal of irradiance, simplifies to the widely-used Jassby & Platt photosynthesis-irradiance curve in the absence of photoinhibition, and only requires one new parameter. This photoinhibition parameter identifies the onset of photoinhibition and is the rate of decrease in photosynthetic rate at that irradiance. The parameters common to both of these models have a consistent interpretation, but our analysis of 3,615 experiments and 10,000 simulated datasets revealed significant discrepancies in parameter interpretation across other widely-used photoinhibition models. Deviations in parameter estimates across models of up to 40% were observed underscoring the need for consistent modeling approaches and supporting the application of our new photoinhibition model.

In this paper, we investigate the impact of speech temporal dynamics in application to automatic speaker verification and speaker voice anonymization tasks. We propose several metrics to perform automatic speaker verification based only on phoneme durations. Experimental results demonstrate that phoneme durations leak some speaker information and can reveal speaker identity from both original and anonymized speech. Thus, this work emphasizes the importance of taking into account the speaker's speech rate and, more importantly, the speaker's phonetic duration characteristics, as well as the need to modify them in order to develop anonymization systems with strong privacy protection capacity., Comment: Accepted at ICASSP 2025