Search

An enhanced version of the conformal BMS$_{3}$ algebra is presented. It is shown to emerge from the asymptotic structure of an extension of conformal gravity in 3D by Pope and Townsend that consistently accommodates an additional spin-2 field, once it is endowed with a suitable set of boundary conditions. The canonical generators of the asymptotic symmetries then span a precise nonlinear W$_{(2,2,2,2,1,1,1)}$ algebra, whose central extensions and coefficients of the nonlinear terms are completely determined by the central charge of the Virasoro subalgebra. The wedge algebra corresponds to the conformal group in four dimensions $SO(4,2)$ and therefore, enhanced conformal BMS$_{3}$ can also be regarded as an infinite-dimensional nonlinear extension of the AdS$_{5}$ algebra with nontrivial central extensions. It is worth mentioning that our boundary conditions might be considered as a starting point in order to consistently incorporate either a finite or an infinite number of conformal higher spin fields., Comment: 10 pages

Quasiequilibrium magnon Bose-Einstein condensates in ferromagnetic insulators have been a field of much interest, while condensation in antiferromagnetic systems has not yet been explored in detail. We analyze the stability of condensed chiral magnons in two antiferromagnetic insulators: a uniaxial easy-axis system and a biaxial system. We show that two-component magnon condensation and inter-magnon interactions are essential to create metastable magnon condensation. The uniaxial system with a Rashba-type Dzyaloshinskii-Moriya interaction supports two degenerate condensate populations at finite wave vectors. We find that the condensation state in this model is stable only when the distribution of condensed magnons between the two populations is symmetric. In addition, we demonstrate the emergence of a zero-sound-like Goldstone mode in antiferromagnetic systems that support two-condensate magnon states. On the other hand, in the biaxial system without Dzyaloshinskii-Moriya interaction, we predict that the magnon condensate cannot stabilize due to the breaking of the magnon degeneracy. Our results suggest that this instability is a general characteristic of single-component quasiequilibrium quasiparticle condensates., Comment: 9 pages, 2 figures

Magnetic-ferroic ordering and magnetic-toroidal moments are essential concepts in molecular electronics and magnetics. The magnetic toroidal moment is critical in understanding new electronic states and their possible uses. This paper discusses the notion of toroidicity waves. In particular, we present a one-dimensional model of interconnected toroidicity leading to an organization principle around an emergent quantum particle, a carrier of toroidicity waves, dubbed the toroidon. We illustrate some functionalities that could be achieved once control over the toroidon is acquired. We show that a 1D dimerized and antiferromagnetic-like spin chain can display ferrotoroidicity and propose its description in terms of an effective quasi-1D spin chain, marking a crucial step towards further research on the phenomena and potential applications of ferrotoroids., Comment: 5 pages, 3 figures

We demonstrate that domain walls built from bimeron chains (bc-DW) in two-dimensional systems constitute a spontaneously assembled medium that holds magnonic excitations along its direction. We prove that such magnons are topological, leading to protected edge states. We also verify the stability of the domain walls and its edge modes' resilience against disorder. Analytical calculations and micromagnetic simulations support our findings. The robustness of these edge modes holds promise for potential applications in the design of nanoscale magnonic devices for information storage and transport., Comment: 5 pages and 5 figures

The aim of this study is to identify quiescent galaxies in the 2-deg$^2$ COSMOS field at $z \sim 3.1$ and analyze their environment. Using data from the ODIN survey and COSMOS2020 catalog, we identify 24 massive quiescent galaxies (MQGs) with stellar masses $\geq 10^{10.6}$ and derive their star formation histories and quenching timescales using SED fitting with BAGPIPES. Voronoi-based density maps trace local and large-scale environments using Lyman-$\alpha$ Emitters and photometric galaxies. Results indicate uniformly short quenching timescales ($<$500 Myr) independent of environmental density, suggesting rapid internal mechanisms such as AGN feedback dominate over environmental factors. MQGs do not correlate with protoclusters or filaments, although some are near gas-rich filaments but show no rejuvenation. These findings suggest quenching at high redshift is driven primarily by internal processes rather than environmental interactions.

The One-hundred-deg$^2$ DECam Imaging in Narrowbands (ODIN) survey is carrying out a systematic search for protoclusters during Cosmic Noon, using Ly$\alpha$-emitting galaxies (LAEs) as tracers. Once completed, ODIN aims to identify hundreds of protoclusters at redshifts of 2.4, 3.1, and 4.5 across seven extragalactic fields, covering a total area of up to 91~deg$^2$. In this work, we report strong clustering of high-redshift protoclusters through the protocluster-LAE cross-correlation function measurements of 150 protocluster candidates at $z~=~2.4$ and 3.1, identified in two ODIN fields with a total area of 13.9 deg$^2$. At $z~=~2.4$ and 3.1, respectively, the inferred protocluster biases are $6.6^{+1.3}_{-1.1}$ and $6.1^{+1.3}_{-1.1}$, corresponding to mean halo masses of $\log \langle M /M_\odot\rangle = 13.53^{+0.21}_{-0.24}$ and $12.96^{+0.28}_{-0.33}$. By the present day, these protoclusters are expected to evolve into virialized galaxy clusters with a mean mass of $\sim$ $10^{14.5}~M_\odot$. By comparing the observed number density of protoclusters to that of halos with the measured clustering strength, we find that our sample is highly complete. Finally, the similar descendant masses derived for our samples at $z=2.4$ and 3.1 assuming that the halo number density remains constant suggest that they represent similar structures observed at different cosmic epochs. As a consequence, any observed differences between the two samples can be understood as redshift evolution. The ODIN protocluster samples will thus provide valuable insights into the cosmic evolution of cluster galaxies., Comment: 12 pages, 5 figures, submitted to ApJ

Humanitarian organizations distribute aid to people affected by armed conflicts or natural disasters. Digitalization has the potential to increase the efficiency and fairness of aid-distribution systems, and recent work by Wang et al. has shown that these benefits are possible without creating privacy harms for aid recipients. However, their work only provides a solution for one particular aid-distribution scenario in which aid recipients receive a pre-defined set of goods. Yet, in many situations it is desirable to enable recipients to decide which items they need at each moment to satisfy their specific needs. We formalize these needs into functional, deployment, security, and privacy requirements, and design a privacy-preserving digital wallet for aid distribution. Our smart-card-based solution enables aid recipients to spend a pre-defined budget at different vendors to obtain the items that they need. We prove our solution's security and privacy properties, and show it is practical at scale.

Spin fluctuations in two-dimensional (2D) ferromagnets in the presence of crystalline lattice dislocations are investigated. We show the existence of topologically protected non-propagative modes that localize at dislocations. These in-gap states, coined as {\it magnonic zero-modes}, are characterized by the $Z_2$ topological invariant that derives from parity symmetry broken induced by sublattice magnetic anisotropy. We uncover that bulk topology existing in the perfect crystal is robust under the influence of lattice defects, which is monitored by the real-space Bott index. It is also revealed that the topology of zero-modes remains unaffected when bulk topology becomes trivial and is remarkably resilient against magnetic disorder. Our findings point to the intriguing relationship between topological lattice defects and the spectrum of topological spin excitations., Comment: 5 pages, 4 figures

We test whether Lyman alpha emitters (LAEs) and Lyman-break galaxies (LBGs) can be good tracers of high-z large-scale structures, using the Horizon Run 5 cosmological hydrodynamical simulation. We identify LAEs using the Ly{\alpha} emission line luminosity and its equivalent width, and LBGs using the broad-band magnitudes at z~2.4, 3.1, and 4.5. We first compare the spatial distributions of LAEs, LBGs, all galaxies, and dark matter around the filamentary structures defined by dark matter. The comparison shows that both LAEs and LBGs are more concentrated toward the dark matter filaments than dark matter. We also find an empirical fitting formula for the vertical density profile of filaments as a binomial power-law relation of the distance to the filaments. We then compare the spatial distributions of the samples around the filaments defined by themselves. LAEs and LBGs are again more concentrated toward their filaments than dark matter. We also find the overall consistency between filamentary structures defined by LAEs, LBGs, and dark matter, with the median spatial offsets that are smaller than the mean separation of the sample. These results support the idea that the LAEs and LBGs could be good tracers of large-scale structures of dark matter at high redshifts., Comment: 20 pages, 15 figures, 2 tables, accepted for publication in ApJ (affiliations are updated in v2)

Differential privacy (DP) is a widely used approach for mitigating privacy risks when training machine learning models on sensitive data. DP mechanisms add noise during training to limit the risk of information leakage. The scale of the added noise is critical, as it determines the trade-off between privacy and utility. The standard practice is to select the noise scale to satisfy a given privacy budget $\varepsilon$. This privacy budget is in turn interpreted in terms of operational attack risks, such as accuracy, sensitivity, and specificity of inference attacks aimed to recover information about the training data records. We show that first calibrating the noise scale to a privacy budget $\varepsilon$, and then translating {\epsilon} to attack risk leads to overly conservative risk assessments and unnecessarily low utility. Instead, we propose methods to directly calibrate the noise scale to a desired attack risk level, bypassing the step of choosing $\varepsilon$. For a given notion of attack risk, our approach significantly decreases noise scale, leading to increased utility at the same level of privacy. We empirically demonstrate that calibrating noise to attack sensitivity/specificity, rather than $\varepsilon$, when training privacy-preserving ML models substantially improves model accuracy for the same risk level. Our work provides a principled and practical way to improve the utility of privacy-preserving ML without compromising on privacy. The code is available at https://github.com/Felipe-Gomez/riskcal, Comment: Appears in NeurIPS 2024

To understand the formation and evolution of massive cosmic structures, studying them at high redshift, in the epoch when they formed the majority of their mass is essential. The One-hundred-deg$^2$ DECam Imaging in Narrowbands (ODIN) survey is undertaking the widest-area narrowband program to date, to use Ly$\alpha$-emitting galaxies (LAEs) to trace the large-scale structure (LSS) of the Universe on the scale of 10 - 100 cMpc at three cosmic epochs. In this work, we present results at $z$ = 3.1 based on early ODIN data in the COSMOS field. We identify and characterize protoclusters and cosmic filaments using multiple methods and discuss their strengths and weaknesses. We then compare our observations against the IllustrisTNG suite of cosmological hydrodynamical simulations. The two are in excellent agreement, with a similar number and angular size of structures identified above a specified density threshold. We are able to recover the simulated protoclusters with $\log$(M$_{z=0}$/$M_\odot$) $\gtrsim$ 14.4 in $\sim$ 60% of the cases. With these objects we show that the descendant masses of the protoclusters in our sample can be estimated purely based on our 2D measurements, finding a median $z$ = 0 mass of $\sim10^{14.5}$M$_\odot$. The lack of information on the radial extent of each protocluster introduces a $\sim$0.4 dex uncertainty in its descendant mass. Finally, we show that the recovery of the cosmic web in the vicinity of protoclusters is both efficient and accurate. The similarity of our observations and the simulations imply that our structure selection is likewise robust and efficient, demonstrating that LAEs are reliable tracers of the LSS., Comment: 29 pages, 20 figures; Accepted for publication in ApJ

Information-based attacks on social media, such as disinformation campaigns and propaganda, are emerging cybersecurity threats. The security community has focused on countering these threats on social media platforms like X and Reddit. However, they also appear in instant-messaging social media platforms such as WhatsApp, Telegram, and Signal. In these platforms information-based attacks primarily happen in groups and channels, requiring manual moderation efforts by channel administrators. We collect, label, and analyze a large dataset of more than 17 million Telegram comments and messages. Our analysis uncovers two independent, coordinated networks that spread pro-Russian and pro-Ukrainian propaganda, garnering replies from real users. We propose a novel mechanism for detecting propaganda that capitalizes on the relationship between legitimate user messages and propaganda replies and is tailored to the information that Telegram makes available to moderators. Our method is faster, cheaper, and has a detection rate (97.6%) 11.6 percentage points higher than human moderators after seeing only one message from an account. It remains effective despite evolving propaganda.

The interplay of spin and lattice fluctuations in two-dimensional magnets without inversion symmetry is investigated. We find a general form for the magnetoelastic coupling between magnons and existing chiral phonons based on the symmetries of the crystalline lattice. We show that in hexagonal lattices, the coupling of magnons and chiral phonons derives from an anisotropic exchange spin model containing topological phases of magnons. Using the Heisenberg-Kitaev-$\Gamma$ model, we show how magnon-polaron edge states with circular polarization arise from this interaction. Our findings exploit the polarization degrees of freedom in spin-lattice systems, thus setting the ground for the transfer of angular momentum between chiral phonons and magnons., Comment: 7 pages, 2 figures

Privacy-enhancing blocking tools based on filter-list rules tend to break legitimate functionality. Filter-list maintainers could benefit from automated breakage detection tools that allow them to proactively fix problematic rules before deploying them to millions of users. We introduce SINBAD, an automated breakage detector that improves the accuracy over the state of the art by 20%, and is the first to detect dynamic breakage and breakage caused by style-oriented filter rules. The success of SINBAD is rooted in three innovations: (1) the use of user-reported breakage issues in forums that enable the creation of a high-quality dataset for training in which only breakage that users perceive as an issue is included; (2) the use of 'web saliency' to automatically identify user-relevant regions of a website on which to prioritize automated interactions aimed at triggering breakage; and (3) the analysis of webpages via subtrees which enables fine-grained identification of problematic filter rules., Comment: 19 pages, 13 figures, Appearing in IEEE S&P 2024

We propose a magnon-based thermal machine in two-dimensional (2D) magnetic insulators. The thermodynamical cycles are engineered by exposing a magnon spin system to thermal baths at different temperatures and tuning the Dzyaloshinskii-Moriya (DM) interaction. We find for the Otto cycle that a thermal gas of magnons converts a fraction of heat into energy in the form of work, where the efficiency is maximized for specific values of DM, reaching the corresponding Carnot efficiency. We witness a positive to negative net work transition during the cycle that marks the onset of a refrigerator-like behavior. The work produced by the magnonic heat engine enhances the magnon chemical potential. The last enables a spin accumulation that might result in the pumping of spin currents at the interfaces of metal-magnet heterostructures. Our work opens new possibilities for the efficient leverage of conventional two-dimensional magnets., Comment: 10 pages. Supplemental material included

The effects of disorder on the robustness of topological magnon states of two-dimensional ferromagnetic skyrmions is investigated. It is diagnosed by evaluating a real space topological invariant, the bosonic Bott index (BI). The disorder simultaneously breaks the axially symmetric shape and the crystalline ordering of the skyrmions array. The corresponding magnonic fluctuations and band spectrum are determined in terms of magnetic field and strength of disorder. We observe the closing of the existing band gaps as the individual skyrmions start to occupy random positions. The analysis reveals that topological states (TSs) persist beyond the perturbative limit when skyrmions reach the glassy phase. In addition, the localization of topologically protected edge states is weakened by the disordered skyrmion structure with increasing localization length. Our findings shed light on the physical understanding of the coexistence of disordered magnetic textures and their topological spin fluctuations., Comment: 6 pages, 4 figures

We introduce a new family of prompt injection attacks, termed Neural Exec. Unlike known attacks that rely on handcrafted strings (e.g., "Ignore previous instructions and..."), we show that it is possible to conceptualize the creation of execution triggers as a differentiable search problem and use learning-based methods to autonomously generate them. Our results demonstrate that a motivated adversary can forge triggers that are not only drastically more effective than current handcrafted ones but also exhibit inherent flexibility in shape, properties, and functionality. In this direction, we show that an attacker can design and generate Neural Execs capable of persisting through multi-stage preprocessing pipelines, such as in the case of Retrieval-Augmented Generation (RAG)-based applications. More critically, our findings show that attackers can produce triggers that deviate markedly in form and shape from any known attack, sidestepping existing blacklist-based detection and sanitation approaches., Comment: v0.2

Collaborative Machine Learning (CML) allows participants to jointly train a machine learning model while keeping their training data private. In many scenarios where CML is seen as the solution to privacy issues, such as health-related applications, safety is also a primary concern. To ensure that CML processes produce models that output correct and reliable decisions \emph{even in the presence of potentially untrusted participants}, researchers propose to use \textit{robust aggregators} to filter out malicious contributions that negatively influence the training process. In this work, we formalize the two prevalent forms of robust aggregators in the literature. We then show that neither can provide the intended protection: either they use distance-based metrics that cannot reliably identify malicious inputs to training; or use metrics based on the behavior of the loss function which create a conflict with the ability of CML participants to learn, i.e., they cannot eliminate the risk of compromise without preventing learning.

The promise of least-privilege learning -- to find feature representations that are useful for a learning task but prevent inference of any sensitive information unrelated to this task -- is highly appealing. However, so far this concept has only been stated informally. It thus remains an open question whether and how we can achieve this goal. In this work, we provide the first formalisation of the least-privilege principle for machine learning and characterise its feasibility. We prove that there is a fundamental trade-off between a representation's utility for a given task and its leakage beyond the intended task: it is not possible to learn representations that have high utility for the intended task but, at the same time prevent inference of any attribute other than the task label itself. This trade-off holds under realistic assumptions on the data distribution and regardless of the technique used to learn the feature mappings that produce these representations. We empirically validate this result for a wide range of learning techniques, model architectures, and datasets.

This research explores the effectiveness of various Machine Learning (ML) models used to predicting solar radiation at the Central Campus of the State Technical University of Quevedo (UTEQ). The data was obtained from a pyranometer, strategically located in a high area of the campus. This instrument continuously recorded solar irradiance data since 2020, offering a comprehensive dataset encompassing various weather conditions and temporal variations. After a correlation analysis, temperature and the time of day were identified as the relevant meteorological variables that influenced the solar irradiance. Different machine learning algorithms such as Linear Regression, K-Nearest Neighbors, Decision Tree, and Gradient Boosting were compared using the evaluation metrics Mean Squared Error (MSE), Root Mean Squared Error (RMSE), Mean Absolute Error (MAE), and the Coefficient of Determination ($R^2$). The study revealed that Gradient Boosting Regressor exhibited superior performance, closely followed by the Random Forest Regressor. These models effectively captured the non-linear patterns in solar radiation, as evidenced by their low MSE and high $R^2$ values. With the aim of assess the performance of our ML models, we developed a web-based tool for the Solar Radiation Forecasting in the UTEQ available at http://https://solarradiationforecastinguteq.streamlit.app/. The results obtained demonstrate the effectiveness of our ML models in solar radiation prediction and contribute a practical utility in real-time solar radiation forecasting, aiding in efficient solar energy management.

Lyman-Alpha Emitting galaxies (LAEs) are typically young, low-mass, star-forming galaxies with little extinction from interstellar dust. Their low dust attenuation allows their Ly$\alpha$ emission to shine brightly in spectroscopic and photometric observations, providing an observational window into the high-redshift universe. Narrowband surveys reveal large, uniform samples of LAEs at specific redshifts that probe large scale structure and the temporal evolution of galaxy properties. The One-hundred-deg$^2$ DECam Imaging in Narrowbands (ODIN) utilizes three custom-made narrowband filters on the Dark Energy Camera (DECam) to discover LAEs at three equally spaced periods in cosmological history. In this paper, we introduce the hybrid-weighted double-broadband continuum estimation technique, which yields improved estimation of Ly$\alpha$ equivalent widths. Using this method, we discover 6032, 5691, and 4066 LAE candidates at $z =$ 2.4, 3.1, and 4.5 in the extended COSMOS field ($\sim$9 deg$^2$). We find that [O II] emitters are a minimal contaminant in our LAE samples, but that interloping Green Pea-like [O III] emitters are important for our redshift 4.5 sample. We introduce an innovative method for identifying [O II] and [O III] emitters via a combination of narrowband excess and galaxy colors, enabling their study as separate classes of objects. We present scaled median stacked SEDs for each galaxy sample, revealing the overall success of our selection methods. We also calculate rest-frame Ly$\alpha$ equivalent widths for our LAE samples and find that the EW distributions are best fit by exponential functions with scale lengths of $w_0$ = 53 $\pm$ 1, 65 $\pm$ 1, and 59 $\pm$ 1 Angstroms, respectively., Comment: Accepted to ApJ, 24 pages, 14 figures

This article primarily aims at classifying, on certain K3 surfaces, the elliptic fibrations induced by conic bundles on smooth del Pezzo surfaces. The key geometric tool employed is the Alexeev-Nikulin correspondence between del Pezzo surfaces with log-terminal singularities of Gorenstein index two and K3 surfaces with non-symplectic involutions of elliptic type: the latter surfaces are realized as appropriate double covers obtained from the former ones. The main application of this correspondence is in the study of linear systems that induce elliptic fibrations on K3 surfaces admitting a strictly elliptic non-symplectic involution, i.e., whose fixed locus consists of a single curve of genus $g\geq 2$. The obtained results are similar to those achieved by Garbagnati and Salgado for jacobian elliptic fibrations., Comment: v2: 21 pages. Corrections in Table 1. Comments are very welcome!

In this paper we introduce a method that supports the detection, identification and localization of lamps in a building, with the main goal of automatically feeding its energy model by means of Building Information Modeling (BIM) methods. The proposed method, thus, provides useful information to apply energy-saving strategies to reduce energy consumption in the building sector through the correct management of the lighting infrastructure. Based on the unique geometry and brightness of lamps and the use of only greyscale images, our methodology is able to obtain accurate results despite its low computational needs, resulting in near-real-time processing. The main novelty is that the focus of the candidate search is not over the entire image but instead only on a limited region that summarizes the specific characteristics of the lamp. The information obtained from our approach was used on the Green Building XML Schema to illustrate the automatic generation of BIM data from the results of the algorithm., Comment: 12 pages, 19 figures, journal

Computer vision is used in this work to detect lighting elements in buildings with the goal of improving the accuracy of previous methods to provide a precise inventory of the location and state of lamps. Using the framework developed in our previous works, we introduce two new modifications to enhance the system: first, a constraint on the orientation of the detected poses in the optimization methods for both the initial and the refined estimates based on the geometric information of the building information modelling (BIM) model; second, an additional reprojection error filtering step to discard the erroneous poses introduced with the orientation restrictions, keeping the identification and localization errors low while greatly increasing the number of detections. These~enhancements are tested in five different case studies with more than 30,000 images, with results showing improvements in the number of detections, the percentage of correct model and state identifications, and the distance between detections and reference positions

This paper introduces a complete method for the automatic detection, identification and localization of lighting elements in buildings, leveraging the available building information modeling (BIM) data of a building and feeding the BIM model with the new collected information, which is key for energy-saving strategies. The detection system is heavily improved from our previous work, with the following two main contributions: (i) a new refinement algorithm to provide a better detection rate and identification performance with comparable computational resources and (ii) a new plane estimation, filtering and projection step to leverage the BIM information earlier for lamps that are both hanging and embedded. The two modifications are thoroughly tested in five different case studies, yielding better results in terms of detection, identification and localization.

Hybrid superconductor(S)-semiconductor(Sm) devices bring a range of new functionalities into superconducting circuits. In particular, hybrid parity-protected qubits and Josephson diodes were recently proposed and experimentally demonstrated. Such devices leverage the non-sinusoidal character of the Josephson current-phase relation (CPR) in highly transparent S-Sm-S junctions. Here we report an experimental study of superconducting quantum-interference devices (SQUIDs) embedding Josephson field-effect transistors fabricated from a SiGe/Ge/SiGe heterostructure grown on a 200-mm silicon wafer. The single-junction CPR shows up to three harmonics with gate tunable amplitude. In the presence of microwave irradiation, the ratio of the first two dominant harmonics, corresponding to single and double Cooper-pair transport processes, is consistently reflected in relative weight of integer and half-integer Shapiro steps. A combination of magnetic-flux and gate-voltage control enables tuning the SQUID functionality from a nonreciprocal Josephson-diode regime with 27% asymmetry to a $\pi$-periodic Josephson regime suitable for the implementation of parity-protected superconducting qubits. These results illustrate the potential of Ge-based hybrid devices as versatile and scalable building blocks of novel superconducting quantum circuits., Comment: 8 pages, 5 figures

The rapid growth of Internet of Things (IoT) devices and applications has led to an increased demand for advanced analytics and machine learning techniques capable of handling the challenges associated with data privacy, security, and scalability. Federated learning (FL) and blockchain technologies have emerged as promising approaches to address these challenges by enabling decentralized, secure, and privacy-preserving model training on distributed data sources. In this paper, we present a novel IoT solution that combines the incremental learning vector quantization algorithm (XuILVQ) with Ethereum blockchain technology to facilitate secure and efficient data sharing, model training, and prototype storage in a distributed environment. Our proposed architecture addresses the shortcomings of existing blockchain-based FL solutions by reducing computational and communication overheads while maintaining data privacy and security. We assess the performance of our system through a series of experiments, showcasing its potential to enhance the accuracy and efficiency of machine learning tasks in IoT settings.