Search

The tokenization of speech with neural audio codec models is a vital part of modern AI pipelines for the generation or understanding of speech, alone or in a multimodal context. Traditionally such tokenization models have concentrated on low parameter-count architectures using only components with strong inductive biases. In this work we show that by scaling a transformer architecture with large parameter count to this problem, and applying a flexible Finite Scalar Quantization (FSQ) based bottleneck, it is possible to reach state-of-the-art speech quality at extremely low bit-rates of $400$ or $700$ bits-per-second. The trained models strongly out-perform existing baselines in both objective and subjective tests.

Theories have revealed the universality of the band tilting effect in topological Weyl semimetals (WSMs) and its implications for the material's physical properties. However, the experimental identification of tilted Weyl bands remains much less explored. Here, by combining magneto-infrared optical studies with a four-band coupled Weyl point model, we report spectroscopic evidence of the tilting effect in the well-established WSM niobium phosphide. Specifically, we observe Landau level transitions with rich features that are well reproduced within a model of coupled tilted Weyl points. Our analysis indicates that the tilting effect relaxes the selection rules and gives rise to transitions that would otherwise be forbidden in the non-tilt case. Additionally, we observe unconventional interband transitions with flat and negative magnetic field dispersions, highlighting the importance of coupling between Weyl points. Our results not only emphasize the significance of the tilting effect in the optical responses of WSMs but also demonstrate magneto-optics as an effective tool for probing the tilting effect in electronic band structures.

Van der Waals (vdW) materials are becoming increasingly popular in scientific and industrial applications because of their unique mixture of record electronic, optical, and mechanical properties. However, nanostructuring of vdW materials is still in its infancy and strongly depends on the specific vdW crystal. As a result, the universal self-assembled technology of vdW materials nanostructuring opens vast technological prospects. This work demonstrates an express and universal synthesis method of vdW nanoparticles with well-defined geometry using femtosecond laser ablation and fragmentation. The disarming simplicity of the technique allows us to create nanoparticles from over 50 vdW precursor materials covering transition metal chalcogenides, MXenes, and other vdW materials. Obtained nanoparticles manifest perfectly defined crystalline structures and diverse shapes, from nanospheres to nanocubes and nanotetrahedrons. Thus, our work provides a new paradigm for vdW nanostructuring with a vast potential of tunability for size, shape, and materials specific to the particular application., Comment: 24 pages, 6 figures

The use of machine learning to predict wave dynamics is a topic of growing interest, but commonly-used deep learning approaches suffer from a lack of interpretability of the trained models. Here we present an interpretable machine learning framework for analyzing the nonlinear evolution dynamics of optical wavepackets in complex wave media. We use sparse regression to reduce microscopic discrete lattice models to simpler effective continuum models which can accurately describe the dynamics of the wavepacket envelope. We apply our approach to valley-Hall domain walls in honeycomb photonic lattices of laser-written waveguides with Kerr-type nonlinearity and different boundary shapes. The reconstructed equations accurately reproduce the linear dispersion and nonlinear effects including self-steepening and self-focusing. This scheme is proven free of the a priori limitations imposed by the underlying hierarchy of scales traditionally employed in asymptotic analytical methods. It represents a powerful interpretable machine learning technique of interest for advancing design capabilities in photonics and framing the complex interaction-driven dynamics in various topological materials., Comment: 6 pages, 5 figures

We study vertical resonant trapping and resonant heating of orbits. These two processes both lead to the growth of a boxy/peanut-shaped bulge in a typical $N$-body model. For the first time, we study this by means of the action variables and resonant angles of the actual orbits that compose the model itself. We used the resonant angle instead of the frequency ratio, which allowed us to clearly distinguish between these two processes in numerical simulations. We show that trapping and heating occur simultaneously, at least at the stage of a mature bar, that is, some orbits quickly pass through vertical resonance while at the same time, a substantial number of orbits remains trapped into this stage for a long time. Half of all bar orbits spend more than 2.5 Gyr in vertical resonance over an interval of 4 Gyr. Half of the orbits trapped into the bar over the last 3 Gyr of simulation remain captured in vertical resonance for more than 2 Gyr. We conclude that in the later stages of the bar evolution, the process of vertical trapping dominates in the ongoing process that causes the boxy/peanut shape of a bar in a typical $N$-body model. This contradicts the results of several recent works., Comment: Accepted for publication in A&A

Traffic signal control is crucial for managing congestion in urban networks. Existing myopic pressure-based control methods focus only on immediate upstream links, leading to suboptimal green time allocation and increased network delays. Effective signal control, however, inherently requires a broader spatial scope, as traffic conditions further upstream can significantly impact traffic at the current location. This paper introduces a novel concept based on the Markov chain theory, namely multi-hop upstream pressure, that generalizes the conventional pressure to account for traffic conditions beyond the immediate upstream links. This farsighted and compact metric informs the deep reinforcement learning agent to preemptively clear the present queues, guiding the agent to optimize signal timings with a broader spatial awareness. Simulations on synthetic and realistic (Toronto) scenarios demonstrate controllers utilizing multi-hop upstream pressure significantly reduce overall network delay by prioritizing traffic movements based on a broader understanding of upstream congestion., Comment: 5 tables, 12 figures. arXiv admin note: text overlap with arXiv:2409.00753

We prove a characterization of F-rationality in terms of tight closure of products of parameter ideals. Our results are inspired by the theory of complete ideals for surfaces and, in particular, the fundamental results of Lipman-Teissier and Cutkosky characterizing rational surface singularities in terms of products of complete ideals, but are valid also in higher dimensions.

Adapting pretrained image-based diffusion models to generate temporally consistent videos has become an impactful generative modeling research direction. Training-free noise-space manipulation has proven to be an effective technique, where the challenge is to preserve the Gaussian white noise distribution while adding in temporal consistency. Recently, Chang et al. (2024) formulated this problem using an integral noise representation with distribution-preserving guarantees, and proposed an upsampling-based algorithm to compute it. However, while their mathematical formulation is advantageous, the algorithm incurs a high computational cost. Through analyzing the limiting-case behavior of their algorithm as the upsampling resolution goes to infinity, we develop an alternative algorithm that, by gathering increments of multiple Brownian bridges, achieves their infinite-resolution accuracy while simultaneously reducing the computational cost by orders of magnitude. We prove and experimentally validate our theoretical claims, and demonstrate our method's effectiveness in real-world applications. We further show that our method readily extends to the 3-dimensional space.

This pilot study explores the application of language models (LMs) to model game event sequences, treating them as a customized natural language. We investigate a popular mobile game, transforming raw event data into textual sequences and pretraining a Longformer model on this data. Our approach captures the rich and nuanced interactions within game sessions, effectively identifying meaningful player segments. The results demonstrate the potential of self-supervised LMs in enhancing game design and personalization without relying on ground-truth labels., Comment: published in Workshop on Customizable NLP at EMNLP 2024

Rare-earth orthoferrites are a promising platform for antiferromagnetic spintronics with a rich variety of terahertz spin and lattice dynamics phenomena. For instance, it has been experimentally demonstrated that the light-driven optical phonons can coherently manipulate macroscopic magnetic states via nonlinear magnetophononic effects. Here using TbFeO$_{3}$ as an example, we reveal the origin of the mode mixing between the LO and TO phonons, which is important for understanding of nonlinear phononics. We performed a comprehensive study of the lattice dynamics of the TbFeO$_{3}$ single crystal by polarized infrared and Raman scattering spectroscopic techniques, and experimentally obtained and carefully analyzed the spectra of anisotropic complex dielectric functions in the far-infrared spectral range. This allowed us to reliably identify the symmetries and parameters of most infrared- and Raman-active phonons. Next, the experimental studies were supplemented by the lattice dynamics calculations which allowed us to propose the normal mode assignments. We reveal that the relation between LO and TO polar phonons is complex and does not strictly follow the ``LO-TO rule'' due to the strong mode mixing. We further analyze how displacements of different ions contribute to phonon modes and reveal that magnetic Fe ions are not involved in Raman-active phonons, thus shedding light on a lack of spin phonon coupling for such phonons. The obtained results establish a solid basis for further in-depth experimental research in the field of nonlinear phononics and magnetophononics in rare-earth orthoferrites., Comment: 23 pages, 15 figures, 6 tables

This paper compares the effectiveness of traditional machine learning methods, encoder-based models, and large language models (LLMs) on the task of detecting depression and anxiety. Five datasets were considered, each differing in format and the method used to define the target pathology class. We tested AutoML models based on linguistic features, several variations of encoder-based Transformers such as BERT, and state-of-the-art LLMs as pathology classification models. The results demonstrated that LLMs outperform traditional methods, particularly on noisy and small datasets where training examples vary significantly in text length and genre. However, psycholinguistic features and encoder-based models can achieve performance comparable to language models when trained on texts from individuals with clinically confirmed depression, highlighting their potential effectiveness in targeted clinical applications.

Photon entanglement is indispensable for optical quantum technologies. Measurement-based optical quantum computing and all-optical quantum networks rely on multiphoton cluster states consisting of indistinguishable entangled photons. A promising method for creating such cluster states on demand is spin-photon entanglement using the spin of a resident charge carrier in a quantum dot, precessing in a weak external magnetic field. In this work, we show theoretically and experimentally that spin-photon entanglement is strongly affected by the hidden anisotropy of quantum dots, which can arise from mechanical stress, shape anisotropy and even specific crystal structure. In the measurements of time-resolved photoluminescence and cross-polarized second-order photon correlation function in a magnetic field, the anisotropy manifests itself in the spin dynamics and, as a consequence, in the spin-photon concurrence. The measured time-filtered spin-photon Bell state fidelity depends strongly on the excitation polarization and reaches an extremely high value of 94% at maximum. We specify the magnetic field and excitation polarization directions that maximize spin-photon entanglement and thereby enhance the fidelity of multiphoton entangled states.

CIZA J2242.8+5301, or the Sausage cluster, is well studied over a range of frequencies. Since its first discovery, a lot of interesting features and unique characteristics have been uncovered. In this work, we report some more new morphological features using the uGMRT band-3 and band-4 data. In the north relic, we observe variation in spectral index profiles across the relic width from the east to west, which may indicate a decrease in downstream cooling rate in that direction. We re-confirm the presence of an additional ~ 930 kpc relic in the north. We classify the filamentary source in the downstream region to be a narrow angle tail (NAT) radio galaxy. The bright arc in the east relic shows sub-structure in the spectral index profile, which may indicate the presence of finer filaments. We further report the presence of a double-strand structure in the east relic similar to the 'Toothbrush' relic. We categorize the bright 'L' shaped structure in the southern relic to be a NAT radio galaxy, as well as trace the actual ~ 1.1 Mpc relic component. We re-confirm the existence of the faint southern extent, measuring the relic length to be ~ 1.8 Mpc. Furthermore, we suggest the southern relic to be a union of individual component relics rather than a single giant filamentary relic. Lastly, based on the morphological symmetry between northern and southern relics, we suggest a schematic shock structure associated with the merger event in an attempt to explain their formation scenario., Comment: 23 pages, 18 figures, 3 tables, submitted to ApJ

We find that the double-threshold voltage-current characteristics in the insulating regime in the ultra-clean two-valley two-dimensional electron system in SiGe/Si/SiGe quantum wells are promoted by perpendicular magnetic fields, persisting to an order of magnitude lower voltages and considerably higher electron densities compared to the zero-field case. This observation indicates the perpendicular-magnetic-field stabilization of the quantum electron solid.

In this paper, we discuss the factorization of the Sudakov form factor on the Coulomb branch of maximally supersymmetric Yang-Mills theory in the near mass-shell limit. We unravel all pinch singularities of this observable making use of the Method of Regions. We find their operator content in terms of matrix elements of Wilson lines on semi-infinite and finite intervals for the jet and ultrasoft functions, respectively. However, naive factorization into these incoherent momentum components is broken at two-loop order by effects subleading in the parameter of dimensional regularization. To save the day, we perform an appropriate twisting of the functions involved as well as simultaneous finite scheme transformation of the 't Hooft coupling. Infrared physics of twisted jet and ultrasoft functions is governed by the octagon anomalous dimension, while the untwisted ultrasoft function possesses infrared evolution driven by an anomalous dimension different from the ubiquitous cusp., Comment: 36 pages, 6 figures

For an isotropic medium, a magnetic field applied in the Voigt geometry affects the optical properties in the second order only, so its effect is much weaker than in the Faraday geometry. In this work, we show that, under resonant excitation well beyond the linear regime, the situation changes drastically: A small magnetic linear anisotropy considerably increases, and an even stronger new quadrupole dichroism emerges. The latter manifests itself as 90{\deg}-periodic azimuthal dependence of transmission and anisotropic rotation of the probe polarization plane. These effects are described microscopically in a toy model and their symmetry analysis is presented. Both are observed experimentally on the D2 resonance of cesium vapor and agree with the theoretical predictions. The large magnitude of the quadrupole dichroism makes it promising for magnetometric applications and for studying the effects of anisotropic bleaching., Comment: 8 pages, 7 figures

The emission and escape of Lyman-$\alpha$ photons from star-forming galaxies is determined through complex interactions between the emitted photons and a galaxy's interstellar and circumgalactic gas, causing Lyman-$\alpha$ emitters (LAEs) to commonly appear not as point sources but in spatially extended halos with complex spectral profiles. We develop a 3D spatial-spectral model of Lyman-$\alpha$ halos (LAHs) to replicate LAH observations in integral field spectroscopic studies, such as those made with VLT/MUSE. The profile of this model is a function of 6 key halo properties: the halo- and compact-source exponential scale lengths ($r_{sH}$ and $r_{sC}$), the halo flux fraction ($f_H$), the compact component ellipticity ($q$), the spectral line width ($\sigma$), and the spectral line skewness parameter ($\gamma$). Placing a series of model LAHs into datacubes reflecting observing conditions in the MUSE UDF-Mosaic survey, we test their detection recoverability and determine that $\sigma$, $r_{sH}$, and $f_H$ are expected to have the most significant effect on the detectability of the overall LAH at a given central wavelength and intrinsic line luminosity. We develop a general selection function model spanning a grid of these halo parameters, and with a sample of 145 UDF-Mosaic LAHs with measured halo properties, we derive completeness-corrected, intrinsic distributions of the values of $\sigma$, $r_{sH}$, and $f_H$ for $3

Large language models (LLMs) have significantly improved code generation, particularly in one-pass code generation. However, most existing approaches focus solely on generating code in a single programming language, overlooking the potential of leveraging the multi-language capabilities of LLMs. LLMs have varying patterns of errors across different languages, suggesting that a more robust approach could be developed by leveraging these multi-language outputs. In this study, we propose Multi-Programming Language Ensemble (MPLE), a novel ensemble-based method that utilizes code generation across multiple programming languages to enhance overall performance. By treating each language-specific code generation process as an individual "weak expert" and effectively integrating their outputs, our method mitigates language-specific errors and biases. This multi-language ensemble strategy leverages the complementary strengths of different programming languages, enabling the model to produce more accurate and robust code. Our approach can be seamlessly integrated with commonly used techniques such as the reflection algorithm and Monte Carlo tree search to improve code generation quality further. Experimental results show that our framework consistently enhances baseline performance by up to 17.92% on existing benchmarks (HumanEval and HumanEval-plus), with a standout result of 96.25% accuracy on the HumanEval benchmark, achieving new state-of-the-art results across various LLM models. The code will be released at https://github.com/NinjaTech-AI/MPLE, Comment: Code available at https://github.com/NinjaTech-AI/MPLE

Perimeter control prevents loss of traffic network capacity due to congestion in urban areas. Homogeneous perimeter control allows all access points to a protected region to have the same maximal permitted inflow. However, homogeneous perimeter control performs poorly when the congestion in the protected region is heterogeneous (e.g., imbalanced demand) since the homogeneous perimeter control does not consider location-specific traffic conditions around the perimeter. When the protected region has spatially heterogeneous congestion, it can often make sense to modulate the perimeter inflow rate to be higher near low-density regions and vice versa for high-density regions. To assist with this modulation, we can leverage the concept of 1-hop traffic pressure to measure intersection-level traffic congestion. However, as we show, 1-hop pressure turns out to be too spatially myopic for perimeter control and hence we formulate multi-hop generalizations of pressure that look ``deeper'' inside the perimeter beyond the entry intersection. In addition, we formulate a simple heterogeneous perimeter control methodology that can leverage this novel multi-hop pressure to redistribute the total permitted inflow provided by the homogeneous perimeter controller. Experimental results show that our heterogeneous perimeter control policies leveraging multi-hop pressure significantly outperform homogeneous perimeter control in scenarios where the origin-destination flows are highly imbalanced with high spatial heterogeneity., Comment: 21 pages main body, 12 figures, journal paper

Superclusters are the largest-scale environments where a number of galaxy clusters interact with each other through minor/major mergers and grow via accretion along cosmic filaments. We focus on the A3528 complex in the core of the Shapley Supercluster. This system includes three clusters, A3528 (composed itself by two sub-clusters, namely A3528N and A3528S), A3532 and A3530, and presents a mildly active dynamical state. We study how minor mergers affect the evolution of radio galaxies and whether they are able to re-accelerate relativistic electrons in the ICM. We used observations from the uGMRT (Band 3, 4 and 5) and MeerKAT (L-band) telescopes to obtain images and spectral index maps over a wide frequency band and spatial resolutions. We compare these data with those from the SRG/eROSITA X-ray telescope. We detect faint diffuse radio emission associated with the radio galaxies. The BCGs in A3528S and A3532 show filaments of diffuse radio emission which extend for $\sim200-400$ kpc out of the radio galaxy. The spectral index of these filaments is extremely steep and almost constant ($\alpha\sim -2, -2.5$). Contrary to the radio tails in A3528N, the spectral properties of these radio filaments are not consistent with standard models of plasma ageing. We also detect roundish diffuse radio emission around the BCG in A3528S which could be classified as a radio mini-halo. The radio tail in this cluster appears longer that in earlier detections, being $\sim300$ kpc long at all frequencies. We linked the presence of extended radio emission in the form of filaments and threads in the A3528 complex with the effect of minor mergers. This is reinforced by the increasing X-ray fluctuations in correspondence with the radio extended emission in A3528S. Despite the less energy involved, our findings support the hypothesis that these events can re-energise plasma originating from radio galaxies., Comment: 28 pages, 23 figures and 5 tables; accepted for publication in Astronomy & Astrophysics

DarkSide-20k (DS-20k) is a dark matter detection experiment under construction at the Laboratori Nazionali del Gran Sasso (LNGS) in Italy. It utilises ~100 t of low radioactivity argon from an underground source (UAr) in its inner detector, with half serving as target in a dual-phase time projection chamber (TPC). The UAr cryogenics system must maintain stable thermodynamic conditions throughout the experiment's lifetime of >10 years. Continuous removal of impurities and radon from the UAr is essential for maximising signal yield and mitigating background. We are developing an efficient and powerful cryogenics system with a gas purification loop with a target circulation rate of 1000 slpm. Central to its design is a condenser operated with liquid nitrogen which is paired with a gas heat exchanger cascade, delivering a combined cooling power of >8 kW. Here we present the design choices in view of the DS-20k requirements, in particular the condenser's working principle and the cooling control, and we show test results obtained with a dedicated benchmarking platform at CERN and LNGS. We find that the thermal efficiency of the recirculation loop, defined in terms of nitrogen consumption per argon flow rate, is 95 % and the pressure in the test cryostat can be maintained within $\pm$(0.1-0.2) mbar. We further detail a 5-day cool-down procedure of the test cryostat, maintaining a cooling rate typically within -2 K/h, as required for the DS-20k inner detector. Additionally, we assess the circuit's flow resistance, and the heat transfer capabilities of two heat exchanger geometries for argon phase change, used to provide gas for recirculation. We conclude by discussing how our findings influence the finalisation of the system design, including necessary modifications to meet requirements and ongoing testing activities., Comment: 45 pages, 24 figures

Quantum computing has tremendous potential to overcome some of the fundamental limitations present in classical information processing. Yet, today's technological limitations in the quality and scaling prevent exploiting its full potential. Quantum computing based on superconducting quantum processing units (QPUs) is among the most promising approaches towards practical quantum advantage. In this article the basic technological approach of IQM Quantum Computers is described covering both the QPU and the rest of the full-stack quantum computer. In particular, the focus is on a 20-qubit quantum computer featuring the Garnet QPU and its architecture, which we will scale up to 150 qubits. We also present QPU and system-level benchmarks, including a median 2-qubit gate fidelity of 99.5% and genuinely entangling all 20 qubits in a Greenberger-Horne-Zeilinger (GHZ) state.

The PHENIX Collaboration has actively pursued a Data and Analysis Preservation program since 2019, the first such dedicated effort at RHIC. A particularly challenging aspect of this endeavor is preservation of complex physics analyses, selected for their scientific importance and the value of the specific techniques developed as a part of the research. For this, we have chosen one of the most impactful PHENIX results, the joint study of direct photons and neutral pions in high-energy d+Au collisions. To ensure reproducibility of this analysis going forward, we partitioned it into self-contained tasks and used a combination of containerization techniques, code management, and robust documentation. We then leveraged REANA (the platform for reproducible analysis developed at CERN) to run the required software. We present our experience based on this example, and outline our future plans for analysis preservation., Comment: Paper submitted to the Proceedings of the Advanced Computing and Analysis Techniques 2024 (ACAT) conference

$\require{mediawiki-texvc}$ We demonstrate that the low symmetry of armchair lateral heterojunction between transition metal dichalcogenide monolayers allows for the mixing of $\mathbf{K}_+$ and $\mathbf{K}_-$ valleys. From the tight binding model we estimate the strength of the valley coupling to be of the order of $0.2$ eV$\cdot{\AA}$ for typical heteropairs. We show that the valley mixing gives rise to the in-plane $g$-factor of localized electrons leading to the spin precession in the in-plane magnetic field. We further study the effects of the valley mixing on the fine structure and dynamics of excitons at type-II lateral heterojunctions. We find that the interplay of the valley mixing and long-range exchange interaction leads to the linear polarization of exciton photoluminescence along the armchair heterojunction with the degree of polarization up to $1/3$ under unpolarized excitation. Application of the in-plane magnetic field of the order of 10$-$100 mT in any direction leads to the depolarization of the photoluminescence., Comment: 10 pages, 5 figures

Broadband impedance-matched Josephson parametric amplifiers are key components for high-fidelity single-shot multi-qubit readout. Nowadays, several types of impedance matched parametric amplifiers have been proposed: the first is an impedance-matched parametric amplifier based on a Klopfenstein taper, and the second is an impedance-matched parametric amplifier based on auxiliary resonators. Here, we present the quantum-limited 3-wave-mixing lumped-element SNAIL parametric amplifier with two-units impedance matching transformer. A two-pole Chebyshev matching network with shunted resonators based on parallel-plate capacitors and superconducting planar coil. Operating in a flux-pumped mode, we experimentally demonstrate an average gain of $15dB$ across a $600MHz$ bandwidth, along with an average saturation power of $-107dBm$ and quantum-limited noise temperature., Comment: 7 pages, 3 figures

The bounds on parameters of the eV and higher scale sterile neutrinos from the $0\nu \beta \beta$ decay have been refined and updated. We present a simple and compact analytic expression for the bound in the $\Delta m^2_{41} - \sin^2 2\theta_{14}$ plane, which includes all relevant parameters. Dependencies of the bound on unknown CP-phases and the type of mass spectrum of light neutrinos (mass ordering and level of degeneracy) are studied in detail. We have computed the bounds using the latest and most stringent data from KamLAND-Zen. The projected constraints from future experiments are estimated. The obtained bounds are confronted with positive indications of the presence of sterile neutrinos as well as with the other existing bounds. The $0\nu \beta \beta$ decay results exclude the regions of parameters implied by BEST and Neutrino-4, and the regions indicated by LSND and MiniBooNE are in conflict with $0\nu \beta \beta$ results combined with $\nu_\mu-$ disappearance bounds., Comment: 6 pages + references, 4 figures

We investigate the effects of dark matter annihilation on objects with masses close to the sub-stellar limit, finding that the minimum mass for stable hydrogen burning is larger than the $\sim0.075 M_\odot $ value predicted in the Standard Model. Below this limit, cooling brown dwarfs evolve into stable dark matter-powered objects that we name dark dwarfs. The timescale of this transition depends on the ambient dark matter density $\rho_{\rm DM}$ and circular velocity $v_{\rm DM}$ but is independent of the dark matter mass. We predict a population of dark dwarfs close to the galactic center, where the dark matter density is expected to be $\rho_{\rm DM}\gtrsim 10^{3}$ GeV/cm$^3$. At larger galactic radii the dark matter density is too low for these objects to have yet formed within the age of the universe. Dark dwarfs retain their initial lithium-7 in mass ranges where brown/red dwarfs would destroy it, providing a method for detecting them., Comment: DC+3JS. 5 pages, 5 figures, results can be reproduced using code provided at https://zenodo.org/records/13141908

We study the radio continuum emission of four galaxies experiencing ram-pressure stripping in four clusters of the Shapley supercluster at redshift z~0.05. Multi-band (235-1367 MHz) radio data, complemented by integral-field spectroscopy, allow us to detect and analyse in detail the non-thermal component both in the galaxy discs and the radio continuum tails. Three galaxies present radio continuum tails which are tens of kiloparsecs long. By deriving the radio spectral index in the inner and outer tails and comparing our findings with the distribution of the extraplanar ionised gas and the results of N-body/hydrodynamical simulations, we demonstrate that these tails are caused by the ram pressure which, together with the ionised gas, sweeps the magnetic field from the galaxy discs. We suggest that the radio continuum emission in these tails can be differently powered by (i) in situ star formation; (ii) relativistic electrons stripped from the disc; (iii) shock excitation or a combination of them. All the ram-pressure stripped galaxies are found in environments where cluster-cluster interactions occurred and/or are ongoing thus strongly supporting the thesis that cluster and group collisions and mergers may locally increase the ram pressure and trigger hydrodynamical interactions between the intracluster medium and the interstellar medium of galaxies., Comment: 18 pages, 14 figures, accepted for publication in MNRAS

The 21 cm transition from neutral Hydrogen promises to be the best observational probe of the Epoch of Reionisation (EoR). This has led to the construction of low-frequency radio interferometric arrays, such as the Hydrogen Epoch of Reionization Array (HERA), aimed at systematically mapping this emission for the first time. Precision calibration, however, is a requirement in 21 cm radio observations. Due to the spatial compactness of HERA, the array is prone to the effects of mutual coupling, which inevitably lead to non-smooth calibration errors that contaminate the data. When unsmooth gains are used in calibration, intrinsically spectrally-smooth foreground emission begins to contaminate the data in a way that can prohibit a clean detection of the cosmological EoR signal. In this paper, we show that the effects of mutual coupling on calibration quality can be reduced by applying custom time-domain filters to the data prior to calibration. We find that more robust calibration solutions are derived when filtering in this way, which reduces the observed foreground power leakage. Specifically, we find a reduction of foreground power leakage by 2 orders of magnitude at k=0.5.

The labor market is undergoing rapid changes, with increasing demands on job seekers and a surge in job openings. Identifying essential skills and competencies from job descriptions is challenging due to varying employer requirements and the omission of key skills. This study addresses these challenges by comparing traditional Named Entity Recognition (NER) methods based on encoders with Large Language Models (LLMs) for extracting skills from Russian job vacancies. Using a labeled dataset of 4,000 job vacancies for training and 1,472 for testing, the performance of both approaches is evaluated. Results indicate that traditional NER models, especially DeepPavlov RuBERT NER tuned, outperform LLMs across various metrics including accuracy, precision, recall, and inference time. The findings suggest that traditional NER models provide more effective and efficient solutions for skill extraction, enhancing job requirement clarity and aiding job seekers in aligning their qualifications with employer expectations. This research contributes to the field of natural language processing (NLP) and its application in the labor market, particularly in non-English contexts.

We propose selective debiasing -- an inference-time safety mechanism that aims to increase the overall quality of models in terms of prediction performance and fairness in the situation when re-training a model is prohibitive. The method is inspired by selective prediction, where some predictions that are considered low quality are discarded at inference time. In our approach, we identify the potentially biased model predictions and, instead of discarding them, we debias them using LEACE -- a post-processing debiasing method. To select problematic predictions, we propose a bias quantification approach based on KL divergence, which achieves better results than standard UQ methods. Experiments with text classification datasets demonstrate that selective debiasing helps to close the performance gap between post-processing methods and at-training and pre-processing debiasing techniques.

We propose an experimental study of adaptive time-stepping methods for efficient modeling of the aggregation-fragmentation kinetics. Precise modeling of this phenomena usually requires utilization of the large systems of nonlinear ordinary differential equations and intensive computations. We concentrate on performance of three explicit Runge-Kutta time-integration methods and provide simulations for two types of problems: finding of equilibrium solutions and simulations for kinetics with periodic solutions. The first class of problems may be analyzed through the relaxation of the solution to the stationary state after large time. In this case, the adaptive time-stepping may help to reach it using big steps reducing cost of the calculations without loss of accuracy. In the second case, the problem becomes numerically unstable at certain points of the phase space and may require tiny steps making the simulations very time-consuming. Adaptive criteria allows to increase the steps for most of points and speedup simulations significantly., Comment: 9 pages, 3 figures, 3 tables

Over the last decade, there has been increasing interest in autonomous driving systems. Reinforcement Learning (RL) shows great promise for training autonomous driving controllers, being able to directly optimize a combination of criteria such as efficiency comfort, and stability. However, RL- based controllers typically offer no safety guarantees, making their readiness for real deployment questionable. In this paper, we propose SECRM-2D (the Safe, Efficient and Comfortable RL- based driving Model with Lane-Changing), an RL autonomous driving controller (both longitudinal and lateral) that balances optimization of efficiency and comfort and follows a fixed route, while being subject to hard analytic safety constraints. The aforementioned safety constraints are derived from the criterion that the follower vehicle must have sufficient headway to be able to avoid a crash if the leader vehicle brakes suddenly. We evaluate SECRM-2D against several learning and non-learning baselines in simulated test scenarios, including freeway driving, exiting, merging, and emergency braking. Our results confirm that representative previously-published RL AV controllers may crash in both training and testing, even if they are optimizing a safety objective. By contrast, our controller SECRM-2D is successful in avoiding crashes during both training and testing, improves over the baselines in measures of efficiency and comfort, and is more faithful in following the prescribed route. In addition, we achieve a good theoretical understanding of the longitudinal steady-state of a collection of SECRM-2D vehicles.

Retrieval-augmented generation (RAG) techniques are widely used today to retrieve and present information in a conversational format. This paper presents a set of enhancements to traditional RAG techniques, focusing on large language models (LLMs) fine-tuned and hosted on AWS Trainium and Inferentia2 AI chips via SageMaker. These chips are characterized by their elasticity, affordability, and efficient performance for AI compute tasks. Besides enabling deployment on these chips, this work aims to improve tool usage, add citation capabilities, and mitigate the risks of hallucinations and unsafe responses due to context bias. We benchmark our RAG system's performance on the Natural Questions and HotPotQA datasets, achieving an accuracy of 62% and 59% respectively, exceeding other models such as DBRX and Mixtral Instruct.

The dual-phase liquid argon time projection chamber is presently one of the leading technologies to search for dark matter particles with masses below 10 GeV/c$^2$. This was demonstrated by the DarkSide-50 experiment with approximately 50 kg of low-radioactivity liquid argon as target material. The next generation experiment DarkSide-20k, currently under construction, will use 1,000 times more argon and is expected to start operation in 2027. Based on the DarkSide-50 experience, here we assess the DarkSide-20k sensitivity to models predicting light dark matter particles, including Weakly Interacting Massive Particles (WIMPs) and sub-GeV/c$^2$ particles interacting with electrons in argon atoms. With one year of data, a sensitivity improvement to dark matter interaction cross-sections by at least one order of magnitude with respect to DarkSide-50 is expected for all these models. A sensitivity to WIMP--nucleon interaction cross-sections below $1\times10^{-42}$ cm$^2$ is achievable for WIMP masses above 800 MeV/c$^2$. With 10 years exposure, the neutrino fog can be reached for WIMP masses around 5 GeV/c$^2$., Comment: submitted to Nature Communications

This study explores the application of genetic algorithms in generating highly nonlinear substitution boxes (S-boxes) for symmetric key cryptography. We present a novel implementation that combines a genetic algorithm with the Walsh-Hadamard Spectrum (WHS) cost function to produce 8x8 S-boxes with a nonlinearity of 104. Our approach achieves performance parity with the best-known methods, requiring an average of 49,399 iterations with a 100% success rate. The study demonstrates significant improvements over earlier genetic algorithm implementations in this field, reducing iteration counts by orders of magnitude. By achieving equivalent performance through a different algorithmic approach, our work expands the toolkit available to cryptographers and highlights the potential of genetic methods in cryptographic primitive generation. The adaptability and parallelization potential of genetic algorithms suggest promising avenues for future research in S-box generation, potentially leading to more robust, efficient, and innovative cryptographic systems. Our findings contribute to the ongoing evolution of symmetric key cryptography, offering new perspectives on optimizing critical components of secure communication systems.

An ultra-light axion with CP violating interactions with a dark sector and CP preserving interactions with the visible sector can act as a novel portal between dark matter and the Standard Model. In such theories, dark matter sources an axion field extending over the entire galaxy, the gradient of which can be searched for with precise spin precession experiments. A reinterpretation of existing co-magnetometer data already constrains theories that are consistent with astrophysical bounds, and near-future experiments will begin probing well-motivated models. The required interactions can arise from a confining hidden sector without necessitating fine-tuning of the axion's mass., Comment: 6 pages, 1 figure; comments welcome!

Due to interactions with dark matter, neutrinos can acquire refractive masses which explain the data from oscillation experiments. We study the effects of relic neutrinos with refractive masses on structure formation in the Universe. In the model with a light fermionic mediator, above the resonance energy, $E_R$, associated with the mass of the mediator, refractive masses have all the properties identical to the usual vacuum masses. Below the resonance, refractive masses decrease with neutrino energy, however, they cannot be used in the same way as usual masses. We study the dispersion relations and group velocities of such neutrinos and their dependence on redshift. We show that in the epoch of structure formation, relic neutrinos were ultrarelativistic and essentially massless particles for $E_R = (10 - 10^5)\,$eV. This allows to reconcile the values of masses extracted from oscillation experiments with the stringent bounds on sum of neutrino masses from cosmological surveys., Comment: 6 pages, 3 figures

We compute the potential-graviton contribution to the scattering amplitude, the radial action, and the scattering angle of two extremal black holes in N = 8 supergravity at the fifth post-Minkowskian order to next-to-leading order in a large mass expansion (first self-force order). Properties of classical unitarity cuts allow us to focus on the integration-by-parts reduction of planar integrals, while nonplanar integrals at this order are obtained from the planar ones by straightforward manipulations. We present all master integrals and solve their associated differential equations necessary to evaluate the classical scattering amplitudes of massive scalar particles at this order in all gravitational theories, in particular in N = 8 supergravity, and in general relativity. Despite the appearance of higher-weight generalized polylogarithms and elliptic functions in the solution to the differential equation for master integrals, the final supergravity answer is remarkably simple and contains only (harmonic) polylogarithmic functions up to weight 2. The systematic analysis of elliptic integrals discussed here, as well as the particular organization of boundary integrals in N = 8 observables are independent of supersymmetry and may have wider applications, including to aspects of collider physics., Comment: 50 pages, 6 figures

We explore the bound neutrons decay into invisible particles (e.g., $n\rightarrow 3 \nu$ or $nn \rightarrow 2 \nu$) in the JUNO liquid scintillator detector. The invisible decay includes two decay modes: $ n \rightarrow { inv} $ and $ nn \rightarrow { inv} $. The invisible decays of $s$-shell neutrons in $^{12}{\rm C}$ will leave a highly excited residual nucleus. Subsequently, some de-excitation modes of the excited residual nuclei can produce a time- and space-correlated triple coincidence signal in the JUNO detector. Based on a full Monte Carlo simulation informed with the latest available data, we estimate all backgrounds, including inverse beta decay events of the reactor antineutrino $\bar{\nu}_e$, natural radioactivity, cosmogenic isotopes and neutral current interactions of atmospheric neutrinos. Pulse shape discrimination and multivariate analysis techniques are employed to further suppress backgrounds. With two years of exposure, JUNO is expected to give an order of magnitude improvement compared to the current best limits. After 10 years of data taking, the JUNO expected sensitivities at a 90% confidence level are $\tau/B( n \rightarrow { inv} ) > 5.0 \times 10^{31} \, {\rm yr}$ and $\tau/B( nn \rightarrow { inv} ) > 1.4 \times 10^{32} \, {\rm yr}$., Comment: 28 pages, 7 figures, 4 tables

The Spin Physics Detector collaboration proposes to install a universal detector in the second interaction point of the NICA collider under construction (JINR, Dubna) to study the spin structure of the proton and deuteron and other spin-related phenomena using a unique possibility to operate with polarized proton and deuteron beams at a collision energy up to 27 GeV and a luminosity up to $10^{32}$ cm$^{-2}$ s$^{-1}$. As the main goal, the experiment aims to provide access to the gluon TMD PDFs in the proton and deuteron, as well as the gluon transversity distribution and tensor PDFs in the deuteron, via the measurement of specific single and double spin asymmetries using different complementary probes such as charmonia, open charm, and prompt photon production processes. Other polarized and unpolarized physics is possible, especially at the first stage of NICA operation with reduced luminosity and collision energy of the proton and ion beams. This document is dedicated exclusively to technical issues of the SPD setup construction.

We aim to constrain the gas density and temperature distributions as well as gas masses in several T Tauri protoplanetary disks located in Taurus. We use the 12CO, 13CO, and C18O (2-1) isotopologue emission observed at 0.9 with the IRAM NOrthern Extended Millimeter Array (NOEMA) as part of the MPG-IRAM Observatory Program PRODIGE (PROtostars and DIsks: Global Evolution PIs: P. Caselli & Th. Henning). Our sample consists of Class II disks with no evidence of strong radial substructures. We use thesedata to constrain the thermal and chemical structure of these disks through theoretical models for gas emission. To fit the combined optically thick and thin CO line data in Fourier space, we developed the DiskCheF code, which includes the parameterized disk physical structure, machine-learning (ML) accelerated chemistry, and the RADMC-3D line radiative transfer module. A key novelty of DiskCheF is the fast and feasible ML-based chemistry trained on the extended grid of the disk physical-chemical models precomputed with the ANDES2 code. This ML approach allows complex chemical kinetics models to be included in a time-consuming disk fitting without the need to run a chemical code. We present a novel approach to incorporate chemistry into disk modeling without the need to explicitly calculate a chemical network every time. Using this new disk modeling tool, we successfully fit the 12CO, 13CO, and C18O (2-1) data from the CI, CY, DL, DM, DN, and IQ Tau disks. The combination of optically thin and optically thick CO lines allows us to simultaneously constrain the disk temperature and mass distribution, and derive the CO-based gas masses. These values are in reasonable agreement with the disk dust masses rescaled by a factor of 100 as well as with other indirect gas measurements., Comment: 25 pages, 19 figures