Search

The neglect of modelling both stellar and nebular emission significantly affects the derived physical properties of galaxies, particularly those with high star formation rates. While this issue has been studied, it has not been established a clear threshold for a significant impact on the estimated physical properties of galaxies due to accounting for both stellar and nebular emission. We analyse galaxies from SDSS-DR7 across a wide range of star-forming activity levels, comparing the results obtained from two spectral fitting tools: FADO (which considers both stellar and nebular continuum) and STARLIGHT (only considers the stellar continuum). A strong linear correlation is found between the rest-frame H$\alpha$ and H$\beta$ equivalent widths (EWs) and the optical nebular contribution, identifying these as reliable tracers. The results show that when the nebular contribution exceeds 8% (corresponding to EW(H$\alpha$)$\simeq$500 \r{A} and EW(H$\beta$)$\simeq$110 \r{A}), there is a significant impact on the estimation of galaxy properties, namely stellar mass, age and metallicity. Our results highlight the importance of taking into account both the stellar and nebular continuum when analysing the optical spectra of star-forming galaxies. In particular, this is a fundamental aspect for galaxies with a rest-frame EW(H$\alpha$)$\gtrsim$500 \r{A} (or the scaled value of 375 \r{A} for pseudo-continuum measures). At low redshifts, this mostly impacts extreme emission line galaxies, while at higher redshifts it becomes a dominant aspect given the higher star-forming activity in the younger Universe. In light of current JWST observations and future instruments designed for high-redshift observations, such as MOONS, this reveals as a critical issue to take into consideration., Comment: 16 pages, 12 figures and 1 table. Accepted for publication in A&A

Reasoning language models (RLMs), also known as Large Reasoning Models (LRMs), such as OpenAI's o1 and o3, DeepSeek-V3, and Alibaba's QwQ, have redefined AI's problem-solving capabilities by extending LLMs with advanced reasoning mechanisms. Yet, their high costs, proprietary nature, and complex architectures - uniquely combining Reinforcement Learning (RL), search heuristics, and LLMs - present accessibility and scalability challenges. To address these, we propose a comprehensive blueprint that organizes RLM components into a modular framework, based on a survey and analysis of all RLM works. This blueprint incorporates diverse reasoning structures (chains, trees, graphs, and nested forms), reasoning strategies (e.g., Monte Carlo Tree Search, Beam Search), RL concepts (policy, value models and others), supervision schemes (Outcome-Based and Process-Based Supervision), and other related concepts (e.g., Test-Time Compute, Retrieval-Augmented Generation, agent tools). We also provide detailed mathematical formulations and algorithmic specifications to simplify RLM implementation. By showing how schemes like LLaMA-Berry, QwQ, Journey Learning, and Graph of Thoughts fit as special cases, we demonstrate the blueprint's versatility and unifying potential. To illustrate its utility, we introduce x1, a modular implementation for rapid RLM prototyping and experimentation. Using x1 and a literature review, we provide key insights, such as multi-phase training for policy and value models, and the importance of familiar training distributions. Finally, we discuss scalable RLM cloud deployments and we outline how RLMs can integrate with a broader LLM ecosystem. Our work demystifies RLM construction, democratizes advanced reasoning capabilities, and fosters innovation, aiming to mitigate the gap between "rich AI" and "poor AI" by lowering barriers to RLM design and experimentation.

The mechanisms of Ly$\alpha$ photon escape are key to understanding galaxy evolution and cosmic reionization, yet remain poorly understood. We investigate the UV-continuum sizes of 23 Ly$\alpha$ emitters (LAEs) at Cosmic Noon ($1.7 < z < 3.3$), extending previous size analyses to include fainter galaxies ($M_{\rm UV} \simeq -14$) using gravitational lensing. Our results show that these LAEs are unusually small for their luminosity, with a mean effective radius ($r_{\rm eff}$) of $170 \pm 140$ pc. They follow a distinct size-luminosity relation, with an intercept at $M_{\rm UV} = -21$ approximately three times smaller than typical star-forming galaxies (SFGs) at similar redshifts. This relation, however, is consistent with that of low-redshift Green Pea galaxies, suggesting that LAEs maintain compact sizes across redshifts. We also find that Ly$\alpha$ equivalent width (EW(Ly$\alpha$)) increases with decreasing $r_{\rm eff}$, confirming previous findings. The small sizes of LAEs lead to high star formation surface densities ($\Sigma$SFR $= 1-600 M_{\sun} \ \rm{yr}^{-1} \ \rm{kpc^{-2}}$), clearly separating them from typical SFGs in the $\Sigma$SFR vs. $r_{\rm eff}$ space. Given that high $\Sigma$SFR is linked to strong galactic outflows, our findings imply that compact morphology plays a key role in Ly$\alpha$ escape, likely facilitated by outflows that clear under-dense channels in the ISM. Thus, these results demonstrate that compact size and high $\Sigma$SFR can help identify Ly$\alpha$-emitters., Comment: 11 pages, 5 figures, submitted to ApJ. Comments welcome

Matrix concentration inequalities and their recently discovered sharp counterparts provide powerful tools to bound the spectrum of random matrices whose entries are linear functions of independent random variables. However, in many applications in theoretical computer science and in other areas one encounters more general random matrix models, called matrix chaoses, whose entries are polynomials of independent random variables. Such models have often been studied on a case-by-case basis using ad-hoc methods that can yield suboptimal dimensional factors. In this paper we provide general matrix concentration inequalities for matrix chaoses, which enable the treatment of such models in a systematic manner. These inequalities are expressed in terms of flattenings of the coefficients of the matrix chaos. We further identify a special family of matrix chaoses of combinatorial type for which the flattening parameters can be computed mechanically by a simple rule. This allows us to provide a unified treatment of and improved bounds for matrix chaoses that arise in a variety of applications, including graph matrices, Khatri-Rao matrices, and matrices that arise in average case analysis of the sum-of-squares hierarchy., Comment: 28 pages

This study investigates the mechanisms of Surveillance Capitalism, focusing on personal data transfer during web navigation and searching. Analyzing network traffic reveals how various entities track and harvest digital footprints. The research reveals specific data types exchanged between users and web services, emphasizing the sophisticated algorithms involved in these processes. We present concrete evidence of data harvesting practices and propose strategies for enhancing data protection and transparency. Our findings highlight the need for robust data protection frameworks and ethical data usage to address privacy concerns in the digital age., Comment: SBBD 2024 - Simp\'osio Brasileiro de Banco de Dados

Microstructural changes in solids, driven by energy flows, do not develop in a static continuous space, such as the space considered in conventional plasticity models. The applied forces create an evolving internal energy landscape, which is constrained by crystallography but has characteristic spatial and temporal scales that form dynamically. To describe this view, we replace a common model for the evolution of dislocation substructure in metals with the evolution of microscopic slips in a combinatorial structure referred to as a polytopal cell complex (PCC). The micro-slips are associated with the 2-cells (faces) of the PCC and are driven by the minimisation of a properly defined Lagrangian. The approach provides a comprehensive statistical and thermodynamic description of plastic flow development. It allows the investigation of energy levels associated with different slip systems and reveals the microscopic mechanisms that result in the phenomenon of strain rate sensitivity.

The Euclid Wide Survey (EWS) is predicted to find approximately 170 000 galaxy-galaxy strong lenses from its lifetime observation of 14 000 deg^2 of the sky. Detecting this many lenses by visual inspection with professional astronomers and citizen scientists alone is infeasible. Machine learning algorithms, particularly convolutional neural networks (CNNs), have been used as an automated method of detecting strong lenses, and have proven fruitful in finding galaxy-galaxy strong lens candidates. We identify the major challenge to be the automatic detection of galaxy-galaxy strong lenses while simultaneously maintaining a low false positive rate. One aim of this research is to have a quantified starting point on the achieved purity and completeness with our current version of CNN-based detection pipelines for the VIS images of EWS. We select all sources with VIS IE < 23 mag from the Euclid Early Release Observation imaging of the Perseus field. We apply a range of CNN architectures to detect strong lenses in these cutouts. All our networks perform extremely well on simulated data sets and their respective validation sets. However, when applied to real Euclid imaging, the highest lens purity is just 11%. Among all our networks, the false positives are typically identifiable by human volunteers as, for example, spiral galaxies, multiple sources, and artefacts, implying that improvements are still possible, perhaps via a second, more interpretable lens selection filtering stage. There is currently no alternative to human classification of CNN-selected lens candidates. Given the expected 10^5 lensing systems in Euclid, this implies 10^6 objects for human classification, which while very large is not in principle intractable and not without precedent., Comment: 22 pages, 11 figures, Euclid consortium paper, A&A submitted

The sharing of mobile network infrastructure has become a key topic with the introduction of 5G due to the high costs of deploying such infrastructures, with neutral host models coupled with features such as network function virtualization (NFV) and network slicing emerging as viable solutions for the challenges in this area. With this in mind, this work presents the design, implementation, and test of a flexible infrastructure-sharing 5G network architecture capable of providing services to any type of client, whether an operator or not. The proposed architecture leverages 5G's network slicing for traffic isolation and compliance with the policies of different clients, with roaming employed for the authentication of users of operator clients. The proposed architecture was implemented and tested in a simulation environment using the UERANSIM and Open5GS open-source tools. Qualitative tests successfully validated the authentication and the traffic isolation features provided by the slices for the two types of clients. Results also demonstrate that the proposed architecture has a positive impact on the performance of the neutral host network infrastructure, achieving 61.8% higher throughput and 96.8% lower packet loss ratio (PLR) in a scenario sharing the infrastructure among four clients and eight users when compared to a single client with all the network resources.

Matrix concentration inequalities, intimately connected to the Non-Commutative Khintchine inequality, have been an important tool in both applied and pure mathematics. We study tensor versions of these inequalities, and establish non-asymptotic inequalities for the $\ell_p$ injective norm of random tensors with correlated entries. In certain regimes of $p$ and the tensor order, our tensor concentration inequalities are nearly optimal in their dimension dependencies. We illustrate our result with applications to problems including structured models of random tensors and matrices, tensor PCA, and connections to lower bounds in coding theory. Our techniques are based on covering number estimates as opposed to operator theoretic tools, which also provide a geometric proof of a weaker version of the Non-Commutative Khintchine inequality, motivated by a question of Talagrand.

The bright radio source, GLEAM J091734-001243 (hereafter GLEAM J0917-0012), was previously selected as a candidate ultra-high redshift (z>5) radio galaxy due to its compact radio size and faint magnitude (K(AB)=22.7). Its redshift was not conclusively determined from follow-up millimetre and near-infrared spectroscopy. Here we present new HST WFC3 G141 grism observations which reveal several emission lines including [NeIII]3867, [NeV]3426 and an extended (~4.8 kpc), [OII]3727 line which confirm a redshift of 3.004+/-0.001. The extended component of the [OII]3727 line is co-spatial with one of two components seen at 2.276 GHz in high resolution (60x20 mas) Long Baseline Array data, reminiscent of the alignments seen in local compact radio galaxies. The BEAGLE stellar mass (~2x10^11 Msun) and radio luminosity (L_500MHz}~10^28 W Hz^-1) put GLEAM J0917-0012 within the distribution of the brightest high-redshift radio galaxies at similar redshifts. However, it is more compact than all of them. Modelling of the radio jet demonstrates that this is a young, ~50 kyr old, but powerful, 10^39 W, compact steep spectrum radio source. The weak constraint on the active galactic nucleus bolometric luminosity from the [NeV]3426 line combined with the modelled jet power tentatively implies a large black hole mass, >10^9 Msun, and a low, advection-dominated accretion rate, an Eddington ratio <0.03. The [NeV]3426/[NeIII]3867 vs [OII]3727/[NeIII]3867 line ratios are most easily explained by radiative shock models with precursor photoionisation. Hence, we infer that the line emission is directly caused by the shocks from the jet and that this radio source is one of the youngest and most powerful known at cosmic noon. We speculate that the star-formation in GLEAM J0917-0012 could be on its way to becoming quenched by the jet., Comment: 16 pages, 7 figures, 4 tables, accepted for publication in PASA

A dedicated gamma-ray burst (GRB) afterglow observing program was performed between 2007 and 2016 with GROND, a seven-channel optical and near-infrared imager at the 2.2m telescope of the Max-Planck Society at ESO/La Silla. In this first of a series of papers, we describe the GRB observing plan, providing first readings of all so far unpublished GRB afterglow measurements and some observing statistics. In total, we observed 514 GRBs with GROND, including 434 Swift-detected GRBs, representing 81\% of the observable Swift sample. For GROND-observations within 30 min of the GRB trigger, the optical/NIR afterglow detection rate is 81\% for long- and 57\% for short-duration GRBs. We report the discovery of ten new GRB afterglows plus one candidate, along with redshift estimates (partly improved) for four GRBs and new host detections for seven GRBs. We identify the (already known) afterglow of GRB 140209A as the sixth GRB exhibiting a 2175 Angstroem dust feature. As a side result, we identified two blazars, with one at a redshift of z=3.8 (in the GRB 131209A field)., Comment: 20 pages, 31 figures, accepted for publication in Astronomy & Astrophysics

Cross-lingual summarization aims to bridge language barriers by summarizing documents in different languages. However, ensuring semantic coherence across languages is an overlooked challenge and can be critical in several contexts. To fill this gap, we introduce multi-target cross-lingual summarization as the task of summarizing a document into multiple target languages while ensuring that the produced summaries are semantically similar. We propose a principled re-ranking approach to this problem and a multi-criteria evaluation protocol to assess semantic coherence across target languages, marking a first step that will hopefully stimulate further research on this problem., Comment: Accepted to EMNLP 2024 (Findings)

The unresolved gamma-ray background (UGRB) is a diffuse gamma-ray emission arising from numerous extragalactic sources below the detection threshold and is an important component of the gamma-ray sky. Studying the UGRB is crucial for understanding high-energy astrophysical processes in the universe and for probing fundamental physics, such as the nature of dark matter. In this work, we forecast the cross-correlation between the UGRB and galaxy catalogs from the Dark Energy Spectroscopic Instrument (DESI) survey. First, we study the expected astrophysical contributions to the UGRB and their cross-correlation with DESI spectroscopic galaxies. Our calculations show that the cross-correlation signal-to-noise ratio is expected to be significant, with the highest value predicted to be 20.6 for DESI luminous red galaxies due to a higher predicted overlap in the redshift distribution with the UGRB. We consider two science cases that the UGRB-spectroscopic galaxies cross-correlation can be applied to: 1) measuring the UGRB flux as a function of redshift, achieving a precision of 10\% in some redshift bins, and 2) searching for annihilating dark matter potentially up to a mass of about 300~GeV, three times higher than the currently strongest constraints. This work underscores the importance of cross correlating the UGRB with cosmic large-scale structure tracers and highlights the multiwavelength approaches to advancing our understanding of high-energy astrophysical phenomena and fundamental physics., Comment: 13 pages, 8 figures, comments are welcome!

We show that the extension map \[ \mathcal{E}_{NS}(f)(z)=\frac{f(x+y)+f(x-y)}{2}+i\frac{f(x+y)-f(x-y)}{2}\mbox{ for all }z=x+iy\in\mathbb{H}\,, \] defined by Norton and Sullivan in '96, is the only locally linear extension map taking bi-Lipschitz functions on $\mathbb{R}$ to quasiconformal functions on $\mathbb{H}$, modulo the action of a group isomorphic to the linear group. In fact, we discovered many other extension like this one (lying in the orbit of such group action), such as: $f(x)\mapsto f(x)+i(f(x)-f(x-y))$.

A vishing attack is a form of social engineering where attackers use phone calls to deceive individuals into disclosing sensitive information, such as personal data, financial information, or security credentials. Attackers exploit the perceived urgency and authenticity of voice communication to manipulate victims, often posing as legitimate entities like banks or tech support. Vishing is a particularly serious threat as it bypasses security controls designed to protect information. In this work, we study the potential for vishing attacks to escalate with the advent of AI. In theory, AI-powered software bots may have the ability to automate these attacks by initiating conversations with potential victims via phone calls and deceiving them into disclosing sensitive information. To validate this thesis, we introduce ViKing, an AI-powered vishing system developed using publicly available AI technology. It relies on a Large Language Model (LLM) as its core cognitive processor to steer conversations with victims, complemented by a pipeline of speech-to-text and text-to-speech modules that facilitate audio-text conversion in phone calls. Through a controlled social experiment involving 240 participants, we discovered that ViKing has successfully persuaded many participants to reveal sensitive information, even those who had been explicitly warned about the risk of vishing campaigns. Interactions with ViKing's bots were generally considered realistic. From these findings, we conclude that tools like ViKing may already be accessible to potential malicious actors, while also serving as an invaluable resource for cyber awareness programs.

Partial least squares (PLS) is a simple factorisation method that works well with high dimensional problems in which the number of observations is limited given the number of independent variables. In this article, we show that PLS can perform better than ordinary least squares (OLS), least absolute shrinkage and selection operator (LASSO) and ridge regression in forecasting quarterly gross domestic product (GDP) growth, covering the period from 2000 to 2023. In fact, through dimension reduction, PLS proved to be effective in lowering the out-of-sample forecasting error, specially since 2020. For the period 2000-2019, the four methods produce similar results, suggesting that PLS is a valid regularisation technique like LASSO or ridge.

We investigate the ability of the Euclid telescope to detect galaxy-scale gravitational lenses. To do so, we perform a systematic visual inspection of the $0.7\,\rm{deg}^2$ Euclid ERO data towards the Perseus cluster using both the high-resolution VIS $I_{\scriptscriptstyle\rm E}$ band, and the lower resolution NISP bands. We inspect every extended source brighter than magnitude $23$ in $I_{\scriptscriptstyle\rm E}$ with $41$ expert human classifiers. This amounts to $12\,086$ stamps of $10^{\prime\prime}\,\times\,10^{\prime\prime}$. We find $3$ grade A and $13$ grade B candidates. We assess the validity of these $16$ candidates by modelling them and checking that they are consistent with a single source lensed by a plausible mass distribution. Five of the candidates pass this check, five others are rejected by the modelling and six are inconclusive. Extrapolating from the five successfully modelled candidates, we infer that the full $14\,000\,{\rm deg}^2$ of the Euclid Wide Survey should contain $100\,000^{+70\,000}_{-30\,000}$ galaxy-galaxy lenses that are both discoverable through visual inspection and have valid lens models. This is consistent with theoretical forecasts of $170\,000$ discoverable galaxy-galaxy lenses in Euclid. Our five modelled lenses have Einstein radii in the range $0.\!\!^{\prime\prime}68\,<\,\theta_\mathrm{E}\,<1.\!\!^{\prime\prime}24$, but their Einstein radius distribution is on the higher side when compared to theoretical forecasts. This suggests that our methodology is likely missing small Einstein radius systems. Whilst it is implausible to visually inspect the full Euclid data set, our results corroborate the promise that Euclid will ultimately deliver a sample of around $10^5$ galaxy-scale lenses., Comment: 21 pages, 20 figures, submitted to A&A

We present deep near-infrared $K_{\rm s}$-band imaging for 35 of the 53 sources from the high-redshift ($z > 2$) radio galaxy candidate sample defined in Broderick et al. (2022). These images were obtained using the High-Acuity Widefield $K$-band Imager (HAWK-I) on the Very Large Telescope. Host galaxies are detected for 27 of the sources, with $K_{\rm s} \approx 21.6$$-$$23.0$ mag (2$''$ diameter apertures; AB). The remaining eight targets are not detected to a median $3\sigma$ depth of $K_{\rm s} \approx 23.3$ mag ($2''$ diameter apertures). We examine the radio and near-infrared flux densities of the 35 sources, comparing them to the known $z > 3$ powerful radio galaxies with 500-MHz radio luminosities $L_{500\,{\rm MHz}} > 10^{27}$ W Hz$^{-1}$. By plotting 150-MHz flux density versus $K_{\rm s}$-band flux density, we find that, similar to the sources from the literature, these new targets have large radio to near-infrared flux density ratios, but extending the distribution to fainter flux densities. Five of the eight HAWK-I deep non-detections have a median $3\sigma$ lower limit of $K_{\rm s} \gtrsim 23.8$ mag ($1.5''$ diameter apertures); these five targets, along with a further source from Broderick et al. (2022) with a deep non-detection ($K_{\rm s} \gtrsim 23.7$ mag; $3\sigma$; $2''$ diameter aperture) in the Southern H-ATLAS Regions $K_{\rm s}$-band Survey, are considered candidates to be ultra-high-redshift ($z > 5$) radio galaxies. The extreme radio to near-infrared flux density ratios ($>10^5$) for these six sources are comparable to TN J0924$-$2201, GLEAM J0856$+$0223 and TGSS J1530$+$1049, the three known powerful radio galaxies at $z > 5$. For a selection of galaxy templates with different stellar masses, we show that $z \gtrsim 4.2$ is a plausible scenario for our ultra-high-redshift candidates if the stellar mass $M_{\rm *} \gtrsim 10^{10.5}\,{\rm M}_\odot$. [abridged], Comment: 24 pages, 5 figures (one of which is a multi-page figure with 30 separate panels), 2 tables, accepted for publication in PASA

Decision Transformer (DT), as one of the representative Reinforcement Learning via Supervised Learning (RvS) methods, has achieved strong performance in offline learning tasks by leveraging the powerful Transformer architecture for sequential decision-making. However, in adversarial environments, these methods can be non-robust, since the return is dependent on the strategies of both the decision-maker and adversary. Training a probabilistic model conditioned on observed return to predict action can fail to generalize, as the trajectories that achieve a return in the dataset might have done so due to a suboptimal behavior adversary. To address this, we propose a worst-case-aware RvS algorithm, the Adversarially Robust Decision Transformer (ARDT), which learns and conditions the policy on in-sample minimax returns-to-go. ARDT aligns the target return with the worst-case return learned through minimax expectile regression, thereby enhancing robustness against powerful test-time adversaries. In experiments conducted on sequential games with full data coverage, ARDT can generate a maximin (Nash Equilibrium) strategy, the solution with the largest adversarial robustness. In large-scale sequential games and continuous adversarial RL environments with partial data coverage, ARDT demonstrates significantly superior robustness to powerful test-time adversaries and attains higher worst-case returns compared to contemporary DT methods., Comment: Accepted to NeurIPS 2024

Observations of the 21-cm signal are opening a window to the cosmic-dawn epoch, when the first stars formed. These observations are usually interpreted with semi-numerical or hydrodynamical simulations, which are often computationally intensive and inflexible to changes in cosmological or astrophysical effects. Here, we present an effective, fully analytic model for the impact of the first stars on the 21-cm signal, using the modular code Zeus21. Zeus21 employs an analytic prescription of the star formation rate density (SFRD) to recover the fully nonlinear and nonlocal correlations of radiative fields that determine the 21-cm signal. We introduce the earliest Population III (Pop III) stars residing in low-mass molecular-cooling galaxies in Zeus21, with distinct spectra from later Pop II stars. We also self-consistently model feedback in the form of $H_2$-dissociating Lyman-Werner (LW) radiation, as well as dark matter-baryon relative velocities, both of which suppress star formation in the lowest-mass halos. LW feedback produces a scale-dependence on the SFRD fluctuations, due to the long mean free path of LW photons. Relative velocities give rise to "wiggles" in the spatial distribution of the 21-cm signal; we present an improved calculation of the shape of these velocity-induced acoustic oscillations, showing they remain a standard ruler at cosmic dawn. Our improved version of Zeus21 predicts the 21-cm global signal and power spectra in agreement with simulations at the $\sim 10\%$ level, yet is at least three orders of magnitude faster. This public code represents a step towards efficient and flexible parameter inference at cosmic dawn, allowing us to predict the first billion years of the universe in mere seconds., Comment: 37 pages, 17 figures, submitted to PRD

SLVideo is a video moment retrieval system for Sign Language videos that incorporates facial expressions, addressing this gap in existing technology. The system extracts embedding representations for the hand and face signs from video frames to capture the signs in their entirety, enabling users to search for a specific sign language video segment with text queries. A collection of eight hours of annotated Portuguese Sign Language videos is used as the dataset, and a CLIP model is used to generate the embeddings. The initial results are promising in a zero-shot setting. In addition, SLVideo incorporates a thesaurus that enables users to search for similar signs to those retrieved, using the video segment embeddings, and also supports the edition and creation of video sign language annotations. Project web page: https://novasearch.github.io/SLVideo/, Comment: 4 pages, 1 figure, 1 table

We study the optimization landscape of a smooth nonconvex program arising from synchronization over the two-element group $\mathbf{Z}_2$, that is, recovering $z_1, \dots, z_n \in \{\pm 1\}$ from (noisy) relative measurements $R_{ij} \approx z_i z_j$. Starting from a max-cut--like combinatorial problem, for integer parameter $r \geq 2$, the nonconvex problem we study can be viewed both as a rank-$r$ Burer--Monteiro factorization of the standard max-cut semidefinite relaxation and as a relaxation of $\{ \pm 1 \}$ to the unit sphere in $\mathbf{R}^r$. First, we present deterministic, non-asymptotic conditions on the measurement graph and noise under which every second-order critical point of the nonconvex problem yields exact recovery of the ground truth. Then, via probabilistic analysis, we obtain asymptotic guarantees for three benchmark problems: (1) synchronization with a complete graph and Gaussian noise, (2) synchronization with an Erd\H{o}s--R\'enyi random graph and Bernoulli noise, and (3) graph clustering under the binary symmetric stochastic block model. In each case, we have, asymptotically as the problem size goes to infinity, a benign nonconvex landscape near a previously-established optimal threshold for exact recovery; we can approach this threshold to arbitrary precision with large enough (but finite) rank parameter $r$. In addition, our results are robust to monotone adversaries.

Bio-inspired soft robots have already shown the ability to handle uncertainty and adapt to unstructured environments. However, their availability is partially restricted by time-consuming, costly and highly supervised design-fabrication processes, often based on resource intensive iterative workflows. Here, we propose an integrated approach targeting the design and fabrication of pneumatic soft actuators in a single casting step. Molds and sacrificial water-soluble hollow cores are printed using fused filament fabrication (FFF). A heated water circuit accelerates the dissolution of the core's material and guarantees its complete removal from the actuator walls, while the actuator's mechanical operability is defined through finite element analysis (FEA). This enables the fabrication of actuators with non-uniform cross sections under minimal supervision, thereby reducing the number of iterations necessary during the design and fabrication processes. Three actuators capable of bending and linear motion were designed, fabricated, integrated and demonstrated as three different bio-inspired soft robots, an earthworm-inspired robot, a four-legged robot, and a robotic gripper. We demonstrate the availability, versatility and effectiveness of the proposed methods, contributing to accelerating the design and fabrication of soft robots. This study represents a step toward increasing the accessibility of soft robots to people at a lower cost.

Traditional Byzantine Fault Tolerance (BFT) state machine replication protocols assume a partial synchrony model, leading to a design where a leader replica drives the protocol and is replaced after a timeout. Recently, we witnessed a surge of asynchronous BFT protocols, which use randomization to remove the need for bounds on message delivery times, making them more resilient to adverse network conditions. However, existing research proposals still fall short of gaining practical adoption, plausibly because they are not able to combine good performance with a simple design that can be readily understood and adopted. In this paper, we present Alea-BFT, a simple and highly efficient asynchronous BFT protocol, which is gaining practical adoption, namely in Ethereum distributed validators. Alea-BFT brings the key design insight from classical protocols of concentrating part of the work on a single designated replica and incorporates this principle in a simple two-stage pipelined design, with an efficient broadcast led by the designated replica, followed by an inexpensive binary agreement. The evaluation of our research prototype implementation and two real-world integrations in cryptocurrency ecosystems shows excellent performance, improving on the fastest protocol (Dumbo-NG) in terms of latency and displaying good performance under faults., Comment: arXiv admin note: substantial text overlap with arXiv:2202.02071

We introduce CYBER-0, the first zero-order optimization algorithm for memory-and-communication efficient Federated Learning, resilient to Byzantine faults. We show through extensive numerical experiments on the MNIST dataset and finetuning RoBERTa-Large that CYBER-0 outperforms state-of-the-art algorithms in terms of communication and memory efficiency while reaching similar accuracy. We provide theoretical guarantees on its convergence for convex loss functions.

The first paper in this series introduced a new family of nonasymptotic matrix concentration inequalities that sharply capture the spectral properties of very general Gaussian (as well as non-Gaussian) random matrices in terms of an associated noncommutative model. These methods achieved matching upper and lower bounds for smooth spectral statistics, but only provided upper bounds for the spectral edges. Here we obtain matching lower bounds for the spectral edges, completing the theory initiated in the first paper. The resulting two-sided bounds enable the study of applications that require an exact determination of the spectral edges to leading order, which is fundamentally beyond the reach of classical matrix concentration inequalities. To illustrate their utility, we undertake a detailed study of phase transition phenomena for spectral outliers of nonhomogeneous random matrices., Comment: 47 pages, 5 figures; part II of series (part I is arXiv:2108.06312)

Generative Artificial Intelligence (GenAI) has become an emerging technology with the availability of several tools that could impact Software Engineering (SE) activities. As any other disruptive technology, GenAI led to the speculation that its full potential can deeply change SE. However, an overfocus on improving activities for which GenAI is more suitable could negligent other relevant areas of the process. In this paper, we aim to explore which SE activities are not expected to be profoundly changed by GenAI. To achieve this goal, we performed a survey with SE practitioners to identify their expectations regarding GenAI in SE, including impacts, challenges, ethical issues, and aspects they do not expect to change. We compared our results with previous roadmaps proposed in SE literature. Our results show that although practitioners expect an increase in productivity, coding, and process quality, they envision that some aspects will not change, such as the need for human expertise, creativity, and project management. Our results point to SE areas for which GenAI is probably not so useful, and future research could tackle them to improve SE practice., Comment: 8 pages, accepted for presentation at the SE2030 - Software Engineering in 2030 Workshop

The AEFishBIT tri-axial accelerometer was externally attached to the operculum to assess the divergent activity and respiratory patterns of two marine farmed fish, the gilthead sea bream (Sparus aurata) and European sea bass (Dicentrarchus labrax). Analysis of raw data from exercised fish highlighted the large amplitude of operculum aperture and body tail movements in European sea bass, which were overall more stable at low-medium exercise intensity levels. Cosinor analysis in free-swimming fish (on-board data processing) highlighted a pronounced daily rhythmicity of locomotor activity and respiratory frequency in both gilthead sea bream and European sea bass. Acrophases of activity and respiration were coupled in gilthead sea bream, acting feeding time (once daily at 11:00 h) as a main synchronizing factor. By contrast, locomotor activity and respiratory frequency were out of phase in European sea bass with activity acrophase on early morning and respiration acrophase on the afternoon. The daily range of activity and respiration variation was also higher in European sea bass, probably as part of the adaptation of this fish species to act as a fast swimming predator. In any case, lower locomotor activity and enhanced respiration were associated with larger body weight in both fish species. This agrees with the notion that selection for fast growth in farming conditions is accompanied by a lower activity profile, which may favor an efficient feed conversion for growth purposes. Therefore, the use of behavioral monitoring is becoming a reliable and large-scale promising tool for selecting more efficient farmed fish, allowing researchers and farmers to establish stricter criteria of welfare for more sustainable and ethical fish production.

We consider a group synchronization problem with multiple frequencies which involves observing pairwise relative measurements of group elements on multiple frequency channels, corrupted by Gaussian noise. We study the computational phase transition in the problem of detecting whether a structured signal is present in such observations by analyzing low-degree polynomial algorithms. We show that, assuming the low-degree conjecture, in synchronization models over arbitrary finite groups as well as over the circle group $SO(2)$, a simple spectral algorithm is optimal among algorithms of runtime $\exp(\tilde{\Omega}(n^{1/3}))$ for detection from an observation including a constant number of frequencies. Combined with an upper bound for the statistical threshold shown in Perry et al., our results indicate the presence of a statistical-to-computational gap in such models with a sufficiently large number of frequencies.

Blockchains revolutionized centralized sectors like banking and finance by promoting decentralization and transparency. In a blockchain, information is transmitted through transactions issued by participants or applications. Miners crucially select, order, and validate pending transactions for block inclusion, prioritizing those with higher incentives or fees. The order in which transactions are included can impact the blockchain final state. Moreover, applications running on top of a blockchain often rely on governance protocols to decentralize the decision-making power to make changes to their core functionality. These changes can affect how participants interact with these applications. Since one token equals one vote, participants holding multiple tokens have a higher voting power to support or reject the proposed changes. The extent to which this voting power is distributed is questionable and if highly concentrated among a few holders can lead to governance attacks. In this thesis, we audit the Bitcoin and Ethereum blockchains to investigate the norms followed by miners in determining the transaction prioritization. We also audit decentralized governance protocols such as Compound to evaluate whether the voting power is fairly distributed among the participants. Our findings have significant implications for future developments of blockchains and decentralized applications., Comment: Doctoral thesis, Saarl\"andische Universit\"ats-und Landesbibliothek, 2023. The original thesis is available at https://publikationen.sulb.uni-saarland.de/handle/20.500.11880/37622

Joint models for longitudinal and survival data have become a popular framework for studying the association between repeatedly measured biomarkers and clinical events. Nevertheless, addressing complex survival data structures, especially handling both recurrent and competing event times within a single model, remains a challenge. This causes important information to be disregarded. Moreover, existing frameworks rely on a Gaussian distribution for continuous markers, which may be unsuitable for bounded biomarkers, resulting in biased estimates of associations. To address these limitations, we propose a Bayesian shared-parameter joint model that simultaneously accommodates multiple (possibly bounded) longitudinal markers, a recurrent event process, and competing risks. We use the beta distribution to model responses bounded within any interval (a,b) without sacrificing the interpretability of the association. The model offers various forms of association, discontinuous risk intervals, and both gap and calendar timescales. A simulation study shows that it outperforms simpler joint models. We utilize the US Cystic Fibrosis Foundation Patient Registry to study the associations between changes in lung function and body mass index, and the risk of recurrent pulmonary exacerbations, while accounting for the competing risks of death and lung transplantation. Our efficient implementation allows fast fitting of the model despite its complexity and the large sample size from this patient registry. Our comprehensive approach provides new insights into cystic fibrosis disease progression by quantifying the relationship between the most important clinical markers and events more precisely than has been possible before. The model implementation is available in the R package JMbayes2.

We aim to identify QSO2 candidates in the redshift desert using optical and infrared photometry. At this intermediate redshift range, most of the prominent optical emission lines in QSO2 sources (e.g. CIV1549; [OIII]4959,5008) fall either outside the wavelength range of the SDSS optical spectra or in particularly noisy wavelength ranges, making QSO2 identification challenging. Therefore, we adopted a semi-supervised machine learning approach to select candidates in the SDSS galaxy sample. Recent applications of machine learning in astronomy focus on problems involving large data sets, with small data sets often being overlooked. We developed a few-shot learning approach for the identification and classification of rare-object classes using limited training data (200 sources). The new AMELIA pipeline uses a transfer-learning based approach with decision trees, distance-based, and deep learning methods to build a classifier capable of identifying rare objects on the basis of an observational training data set. We validated the performance of AMELIA by addressing the problem of identifying QSO2s at 1 $\leq$ z $\leq$ 2 using SDSS and WISE photometry, obtaining an F1-score above 0.8 in a supervised approach. We then used AMELIA to select new QSO2 candidates in the redshift desert and examined the nature of the candidates using SDSS spectra, when available. In particular, we identified a sub-population of [NeV]3426 emitters at z $\sim$ 1.1, which are highly likely to contain obscured AGNs. We used X-ray and radio cross-matching to validate our classification and investigated the performance of photometric criteria from the literature showing that our candidates have an inherent dusty nature. Finally, we derived physical properties for our QSO2 sample using photoionisation models and verified the AGN classification using an SED fitting., Comment: 20 pages, 9 figures, Accepted for publication in A&A

The current standard model of cosmology successfully describes a variety of measurements, but the nature of its main ingredients, dark matter and dark energy, remains unknown. Euclid is a medium-class mission in the Cosmic Vision 2015-2025 programme of the European Space Agency (ESA) that will provide high-resolution optical imaging, as well as near-infrared imaging and spectroscopy, over about 14,000 deg^2 of extragalactic sky. In addition to accurate weak lensing and clustering measurements that probe structure formation over half of the age of the Universe, its primary probes for cosmology, these exquisite data will enable a wide range of science. This paper provides a high-level overview of the mission, summarising the survey characteristics, the various data-processing steps, and data products. We also highlight the main science objectives and expected performance., Comment: Accepted for publication in the A&A special issue`Euclid on Sky'

Geometrically frustrated magnetic systems, such as those based on the Shastry-Sutherland lattice, offer a rich playground for exploring unconventional magnetic states. The delicate balance between competing interactions in these systems leads to the emergence of novel phases. We present the characterization of Er2Be2GeO7, an SSL compound with Er3+ ions forming orthogonal dimers separated by non-magnetic layers. Neutron scattering reveals an antiferromagnetic dimer structure at zero field, typical of Ising spins on that lattice and consistent with the anisotropic magnetization observed. However, magnetization measurements exhibit fractional plateaus at 1/4 and 1/2 of saturation, by contrast to the expected 1/3 plateau of the SSL Ising model. We show that this behaviour requires spatially anisotropic interactions, and, using a tensor network approach, we show that all properties, including thermodynamics, can be quantitatively explained by an Anisotropic Shastry-Sutherland Ising Model whose microscopic origin can be traced back to Dzyaloshinskii-Moriya interactions. This study highlights Er2Be2GeO7 as a promising platform for investigating exotic magnetic phenomena and potential applications in spintronics.

We prove the existence of a positive semidefinite matrix $A \in \mathbb{R}^{n \times n}$ such that any decomposition into rank-1 matrices has to have factors with a large $\ell^1-$norm, more precisely $$ \sum_{k} x_k x_k^*=A \quad \implies \quad \sum_k \|x_k\|^2_{1} \geq c \sqrt{n} \|A\|_{1},$$ where $c$ is independent of $n$. This provides a lower bound for the Balan--Jiang matrix problem. The construction is probabilistic.