Search

The vertical evolution of galactic discs is governed by the sub-structures within them. We examine the diversity of kinematic sub-structure present in the first 12 galaxies observed from the GECKOS survey, a VLT/MUSE large programme providing a systematic study of 35 edge-on, Milky Way-mass disc galaxies. Employing the nGIST analysis pipeline, we derive the mean line-of-sight stellar velocity ($V_{\star}$), velocity dispersion ($\sigma_{\star}$), skew ($h_{3}$), and kurtosis ($h_{4}$) for the sample, and examine 2D maps and 1D line profiles. Visually, the majority of this sample (8/12) are found to possess boxy-peanut bulges and host the corresponding kinematic structure predicted for stellar bars viewed in projection. Four galaxies exhibit strong evidence for the presence of nuclear discs, including central $h_{3}$-$V_{\star}$ anti-correlations, `croissant'-shaped central depressions in $\sigma_{\star}$ maps, strong gradients in $h_{3}$, and positive $h_{4}$ plateaus over the expected nuclear disc extent. The strength of the $h_{3}$ feature corresponds to the size of the nuclear disc, measured from the $h_{3}$ turnover radius. We can explain the features within the kinematic maps of all sample galaxies via disc structure(s) alone. We do not find any need to invoke the existence of dispersion-dominated bulges. Obtaining the specialised data products for this paper and the broader GECKOS survey required significant development of existing integral field spectroscopic (IFS) analysis tools. Therefore, we also present the nGIST pipeline: a modern, sophisticated, and easy-to-use pipeline for the analysis of galaxy IFS data. We conclude that the variety of kinematic sub-structures seen in GECKOS galaxies requires a contemporary view of galaxy morphology, expanding on the traditional view of galaxy structure, and uniting the kinematic complexity observed in the Milky Way with the extragalactic., Comment: 32 pages (9 of which are appendix), 26 figures, submitted to A&A. Comments welcome!

Dengue is a climate-sensitive mosquito-borne disease with a complex transmission dynamic. Data related to climate, environmental and sociodemographic characteristics of the target population are important for project scenarios. Different datasets and methodologies have been applied to build complex models for dengue forecast, stressing the need to evaluate these models and their relative accuracy grounded on a reproducible methodology. The goal of this work is to describe and present Mosqlimate, a web-based platform composed by a dashboard, a data store, model and rediction registries and support for a community of practice in arbovirus forecasting. Multiple API endpoints give access to data for development, open registration of predictive models from different approaches and sharing of predictive models for arboviruses incidence, facilitating interaction between modellers and allowing for proper comparison of the performance of different registered models, by means of probabilistic scores. Epidemiological, entomological, climatic and sociodemographic datasets related to arboviruses in Brazil, are freely available for download, alongside full documentation., Comment: 10 pages, 2 figures, 6 tables

Binary population synthesis (BPS) is an essential tool for extracting information about massive binary evolution from gravitational-wave (GW) detections of compact object mergers. It has been successfully used to constrain the most likely permutations of evolution models among hundreds of alternatives, while initial condition models, in contrast, have not yet received the same level of attention. Here, we introduce BOSSA, a detailed initial sampling code including a set of 192 initial condition permutations for BPS that capture both "invariant" and "varying" models, the latter accounting for a possible metallicity- and star formation rate (SFR)-dependence of the initial mass function (IMF); as well as correlations between the initial primary mass, orbital period, mass ratio and eccentricity of binaries. We include 24 metallicity-specific cosmic star formation history (cSFH) models and propose two alternate models for the mass-dependent binary fraction. We build a detailed pipeline for time-evolving BPS, such that each binary has well-defined initial conditions, and we are able to distinguish the contributions from populations of different ages. We discuss the meaning of the IMF for binaries and introduce a refined initial sampling procedure for component masses. We also discuss the treatment of higher-order multiple systems when normalizing a binary sample. In particular, we argue for how a consistent interpretation of the IMF implies that this is not the distribution from which any set of component masses should be independently drawn, and show how the individual IMF of primaries and companions is expected to deviate from the full IMF., Comment: 23 pages, 16 figures, 2 tables. Submitted to MNRAS. Repository (under construction) at https://github.com/lmdesa/BOSSA

We perform a first study of the impact of varying two components of the initial conditions in binary population synthesis of compact binary mergers - the initial mass function, which is made metallicity- and star formation rate-dependent, and the orbital parameter (orbital period, mass ratio and eccentricity) distributions, which are assumed to be correlated - within a larger grid of initial condition models also including alternatives for the primary mass-dependent binary fraction and the metallicity-specific cosmic star formation history. We generate the initial populations with the sampling code BOSSA and evolve them with the rapid population synthesis code COMPAS. We find strong suggestions that the main role of initial conditions models is to set the relative weights of key features defined by the evolution models. In the two models we compare, black hole-black hole (BHBH) mergers are the most strongly affected, which we connect to a shift from the common envelope to the stable Roche lobe overflow formation channels with decreasing redshift. We also characterize variations in the black hole-neutron star (BHNS) and neutron star-neutron star (NSNS) final parameter distributions. We obtain the merger rate evolution for BHBH, BHNS and NSNS mergers up to $z=10$, and find a variation by a factor of $\sim50-60$ in the local BHBH and BHNS merger rates, suggesting a more important contribution from initial conditions than previously thought, and calling for a complete exploration of the initial conditions model permutations., Comment: 28 pages, 14 figures, 3 tables. Accepted for publication in MNRAS. Code/data to be made publicly available

Dense linear layers are the dominant computational bottleneck in large neural networks, presenting a critical need for more efficient alternatives. Previous efforts focused on a small number of hand-crafted structured matrices and neglected to investigate whether these structures can surpass dense layers in terms of compute-optimal scaling laws when both the model size and training examples are optimally allocated. In this work, we present a unifying framework that enables searching among all linear operators expressible via an Einstein summation. This framework encompasses many previously proposed structures, such as low-rank, Kronecker, Tensor-Train, Block Tensor-Train (BTT), and Monarch, along with many novel structures. To analyze the framework, we develop a taxonomy of all such operators based on their computational and algebraic properties and show that differences in the compute-optimal scaling laws are mostly governed by a small number of variables that we introduce. Namely, a small $\omega$ (which measures parameter sharing) and large $\psi$ (which measures the rank) reliably led to better scaling laws. Guided by the insight that full-rank structures that maximize parameters per unit of compute perform the best, we propose BTT-MoE, a novel Mixture-of-Experts (MoE) architecture obtained by sparsifying computation in the BTT structure. In contrast to the standard sparse MoE for each entire feed-forward network, BTT-MoE learns an MoE in every single linear layer of the model, including the projection matrices in the attention blocks. We find BTT-MoE provides a substantial compute-efficiency gain over dense layers and standard MoE., Comment: NeurIPS 2024. Code available at https://github.com/AndPotap/einsum-search

Automatic facial action unit (AU) recognition is used widely in facial expression analysis. Most existing AU recognition systems aim for cross-participant non-calibrated generalization (NCG) to unseen faces without further calibration. However, due to the diversity of facial attributes across different identities, accurately inferring AU activation from single images of an unseen face is sometimes infeasible, even for human experts -- it is crucial to first understand how the face appears in its neutral expression, or significant bias may be incurred. Therefore, we propose to perform one-frame calibration (OFC) in AU recognition: for each face, a single image of its neutral expression is used as the reference image for calibration. With this strategy, we develop a Calibrating Siamese Network (CSN) for AU recognition and demonstrate its remarkable effectiveness with a simple iResNet-50 (IR50) backbone. On the DISFA, DISFA+, and UNBC-McMaster datasets, we show that our OFC CSN-IR50 model (a) substantially improves the performance of IR50 by mitigating facial attribute biases (including biases due to wrinkles, eyebrow positions, facial hair, etc.), (b) substantially outperforms the naive OFC method of baseline subtraction as well as (c) a fine-tuned version of this naive OFC method, and (d) also outperforms state-of-the-art NCG models for both AU intensity estimation and AU detection.

Machine learning (ML) is revolutionising drug discovery by expediting the prediction of small molecule properties essential for developing new drugs. These properties -- including absorption, distribution, metabolism and excretion (ADME)-- are crucial in the early stages of drug development since they provide an understanding of the course of the drug in the organism, i.e., the drug's pharmacokinetics. However, existing methods lack personalisation and rely on manually crafted ML algorithms or pipelines, which can introduce inefficiencies and biases into the process. To address these challenges, we propose a novel evolutionary-based automated ML method (AutoML) specifically designed for predicting small molecule properties, with a particular focus on pharmacokinetics. Leveraging the advantages of grammar-based genetic programming, our AutoML method streamlines the process by automatically selecting algorithms and designing predictive pipelines tailored to the particular characteristics of input molecular data. Results demonstrate AutoML's effectiveness in selecting diverse ML algorithms, resulting in comparable or even improved predictive performances compared to conventional approaches. By offering personalised ML-driven pipelines, our method promises to enhance small molecule research in drug discovery, providing researchers with a valuable tool for accelerating the development of novel therapeutic drugs., Comment: Paper accepted and presented at the 14th Workshop on Evolutionary Computation for the Automated Design of Algorithms (ECADA), which happened during the Genetic and Evolutionary Computation Conference (GECCO)

Most complex diseases, including cancer and non-malignant diseases like asthma, have distinct molecular subtypes that require distinct clinical approaches. However, existing computational patient stratification methods have been benchmarked almost exclusively on cancer omics data and only perform well when mutually exclusive subtypes can be characterized by many biomarkers. Here, we contribute with a massive evaluation attempt, quantitatively exploring the power of 22 unsupervised patient stratification methods using both, simulated and real transcriptome data. From this experience, we developed UnPaSt (https://apps.cosy.bio/unpast/) optimizing unsupervised patient stratification, working even with only a limited number of subtype-predictive biomarkers. We evaluated all 23 methods on real-world breast cancer and asthma transcriptomics data. Although many methods reliably detected major breast cancer subtypes, only few identified Th2-high asthma, and UnPaSt significantly outperformed its closest competitors in both test datasets. Essentially, we showed that UnPaSt can detect many biologically insightful and reproducible patterns in omic datasets., Comment: The first two authors listed are joint first authors. The last two authors listed are joint last authors

Languages continually evolve in response to societal events, resulting in new terms and shifts in meanings. These changes have significant implications for computer applications, including automatic translation and chatbots, making it essential to characterize them accurately. The recent development of LLMs has notably advanced natural language understanding, particularly in sense inference and reasoning. In this paper, we investigate the potential of LLMs in characterizing three types of semantic change: dimension, relation, and orientation. We achieve this by combining LLMs' Chain-of-Thought with rhetorical devices and conducting an experimental assessment of our approach using newly created datasets. Our results highlight the effectiveness of LLMs in capturing and analyzing semantic changes, providing valuable insights to improve computational linguistic applications.

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.

Post-training quantization (PTQ) reduces the memory footprint of LLMs by quantizing weights to low-precision datatypes. Since LLM inference is usually memory-bound, PTQ methods can improve inference throughput. Recent state-of-the-art PTQ approaches use vector quantization (VQ) to quantize multiple weights at once, which improves information utilization through better shaping. However, VQ requires a codebook with size exponential in the dimension. This limits current VQ-based PTQ works to low VQ dimensions ($\le 8$) that in turn limit quantization quality. Here, we introduce QTIP, which instead uses trellis coded quantization (TCQ) to achieve ultra-high-dimensional quantization. TCQ uses a stateful decoder that separates the codebook size from the bitrate and effective dimension. QTIP introduces a spectrum of lookup-only to computed lookup-free trellis codes designed for a hardware-efficient "bitshift" trellis structure; these codes achieve state-of-the-art results in both quantization quality and inference speed., Comment: NeurIPS 2024 Spotlight

A novel method to reconstruct the energy of hadronic showers in the CMS High Granularity Calorimeter (HGCAL) is presented. The HGCAL is a sampling calorimeter with very fine transverse and longitudinal granularity. The active media are silicon sensors and scintillator tiles readout by SiPMs and the absorbers are a combination of lead and Cu/CuW in the electromagnetic section, and steel in the hadronic section. The shower reconstruction method is based on graph neural networks and it makes use of a dynamic reduction network architecture. It is shown that the algorithm is able to capture and mitigate the main effects that normally hinder the reconstruction of hadronic showers using classical reconstruction methods, by compensating for fluctuations in the multiplicity, energy, and spatial distributions of the shower's constituents. The performance of the algorithm is evaluated using test beam data collected in 2018 prototype of the CMS HGCAL accompanied by a section of the CALICE AHCAL prototype. The capability of the method to mitigate the impact of energy leakage from the calorimeter is also demonstrated., Comment: Prepared for submission to JINST

In this paper we develop an intrinsic formalism to study the topology, smooth structure, and Riemannian geometry of the Wasserstein space of a closed Riemannian manifold. Our formalism allows for a new characterisation of the Weak topology via convergent sequences of the subjacent space. Applying it we also provide a new proof that Wasserstein spaces of closed manifolds are geodesically convex. Our framework is particularly handy to address the Wasserstein spaces of compact Lie groups, where we refine our formalism and present an explicit example., Comment: 24 pages

We study gradient descent (GD) dynamics on logistic regression problems with large, constant step sizes. For linearly-separable data, it is known that GD converges to the minimizer with arbitrarily large step sizes, a property which no longer holds when the problem is not separable. In fact, the behaviour can be much more complex -- a sequence of period-doubling bifurcations begins at the critical step size $2/\lambda$, where $\lambda$ is the largest eigenvalue of the Hessian at the solution. Using a smaller-than-critical step size guarantees convergence if initialized nearby the solution: but does this suffice globally? In one dimension, we show that a step size less than $1/\lambda$ suffices for global convergence. However, for all step sizes between $1/\lambda$ and the critical step size $2/\lambda$, one can construct a dataset such that GD converges to a stable cycle. In higher dimensions, this is actually possible even for step sizes less than $1/\lambda$. Our results show that although local convergence is guaranteed for all step sizes less than the critical step size, global convergence is not, and GD may instead converge to a cycle depending on the initialization.

Zeroth-order optimization (ZO) is a memory-efficient strategy for fine-tuning Large Language Models using only forward passes. However, the application of ZO fine-tuning in memory-constrained settings such as mobile phones and laptops is still challenging since full precision forward passes are infeasible. In this study, we address this limitation by integrating sparsity and quantization into ZO fine-tuning of LLMs. Specifically, we investigate the feasibility of fine-tuning an extremely small subset of LLM parameters using ZO. This approach allows the majority of un-tuned parameters to be quantized to accommodate the constraint of limited device memory. Our findings reveal that the pre-training process can identify a set of "sensitive parameters" that can guide the ZO fine-tuning of LLMs on downstream tasks. Our results demonstrate that fine-tuning 0.1% sensitive parameters in the LLM with ZO can outperform the full ZO fine-tuning performance, while offering wall-clock time speedup. Additionally, we show that ZO fine-tuning targeting these 0.1% sensitive parameters, combined with 4 bit quantization, enables efficient ZO fine-tuning of an Llama2-7B model on a GPU device with less than 8 GiB of memory and notably reduced latency.

We investigate the static boson star solutions in the so-called Einstein-Friedberg-Lee-Sirlin (E-FLS) theory, performing a complete analysis of the solution space in this model. We study the phenomenological aspects of E-FLS stars, for instance, by investigating the timelike and null geodesics with an emphasis on the analysis of circular timelike orbits and light rings. In order to study the astrophysical signatures of such stars, their images were obtained considering them surrounded by a geometrically thin accretion disk. Our results comprise two different models of accretion disks, namely the optically thin and thick disk models. We present a selection of our findings for the astrophysical images of E-FLS stars and discuss their relevance as a possible black hole mimicker., Comment: 16 pages, 13 figures

We present STAT: a simple algorithm to prune transformer models without any fine-tuning. STAT eliminates both attention heads and neurons from the network, while preserving accuracy by calculating a correction to the weights of the next layer. Each layer block in the network is compressed using a series of principled matrix factorizations that preserve the network structure. Our entire algorithm takes minutes to compress BERT, and less than three hours to compress models with 7B parameters using a single GPU. Using only several hundred data examples, STAT preserves the output of the network and improves upon existing gradient-free pruning methods. It is even competitive with methods that include significant fine-tuning. We demonstrate our method on both encoder and decoder architectures, including BERT, DistilBERT, and Llama-2 using benchmarks such as GLUE, Squad, WikiText2.

Background: Timely prioritising and remediating vulnerabilities are paramount in the dynamic cybersecurity field, and one of the most widely used vulnerability scoring systems (CVSS) does not address the increasing likelihood of emerging an exploit code. Aims: We present SecScore, an innovative vulnerability severity score that enhances CVSS Threat metric group with statistical models from empirical evidences of real-world exploit codes. Method: SecScore adjusts the traditional CVSS score using an explainable and empirical method that more accurately and promptly captures the dynamics of exploit code development. Results: Our approach can integrate seamlessly into the assessment/prioritisation stage of several vulnerability management processes, improving the effectiveness of prioritisation and ensuring timely remediation. We provide real-world statistical analysis and models for a wide range of vulnerability types and platforms, demonstrating that SecScore is flexible according to the vulnerability's profile. Comprehensive experiments validate the value and timeliness of SecScore in vulnerability prioritisation. Conclusions: SecScore advances the vulnerability metrics theory and enhances organisational cybersecurity with practical insights.

With the proliferation of Internet of Things (IoT) devices, ensuring secure communications has become imperative. Due to their low cost and embedded nature, many of these devices operate with computational and energy constraints, neglecting the potential security vulnerabilities that they may bring. This work-in-progress is focused on designing secure communication among remote servers and embedded IoT devices to balance security robustness and energy efficiency. The proposed approach uses lightweight cryptography, optimizing device performance and security without overburdening their limited resources. Our architecture stands out for integrating Edge servers and a central Name Server, allowing secure and decentralized authentication and efficient connection transitions between different Edge servers. This architecture enhances the scalability of the IoT network and reduces the load on each server, distributing the responsibility for authentication and key management., Comment: 6 pages and 2 figures

In this work, we have adopted the diffusion-based image reconstruction pipeline previously used for fMRI image reconstruction and applied it to Electroencephalography (EEG). The EEG encoding method is very simple, and forms a baseline from which more sophisticated EEG encoding methods can be compared. We have also evaluated the fidelity of the generated image using the same metrics used in the previous functional magnetic resonance imaging (fMRI) and magnetoencephalography (MEG) works. Our results show that while the reconstruction from EEG recorded to rapidly presented images is not as good as reconstructions from fMRI to slower presented images, it holds a surprising amount of information that could be applied in specific use cases. Also, EEG-based image reconstruction works better in some categories-such as land animals and food-than others, shedding new light on previous findings of EEG's sensitivity to those categories and revealing potential for these methods to further understand EEG responses to human visual coding. More investigation should use longer-duration image stimulations to elucidate the later components that might be salient to the different image categories.

Live languages continuously evolve to integrate the cultural change of human societies. This evolution manifests through neologisms (new words) or \textbf{semantic changes} of words (new meaning to existing words). Understanding the meaning of words is vital for interpreting texts coming from different cultures (regionalism or slang), domains (e.g., technical terms), or periods. In computer science, these words are relevant to computational linguistics algorithms such as translation, information retrieval, question answering, etc. Semantic changes can potentially impact the quality of the outcomes of these algorithms. Therefore, it is important to understand and characterize these changes formally. The study of this impact is a recent problem that has attracted the attention of the computational linguistics community. Several approaches propose methods to detect semantic changes with good precision, but more effort is needed to characterize how the meaning of words changes and to reason about how to reduce the impact of semantic change. This survey provides an understandable overview of existing approaches to the \textit{characterization of semantic changes} and also formally defines three classes of characterizations: if the meaning of a word becomes more general or narrow (change in dimension) if the word is used in a more pejorative or positive/ameliorated sense (change in orientation), and if there is a trend to use the word in a, for instance, metaphoric or metonymic context (change in relation). We summarized the main aspects of the selected publications in a table and discussed the needs and trends in the research activities on semantic change characterization.

We address in this work the nature and evolution of the long-period compact star sources, which has recently added several unexpected members. The central hypothesis is that particle winds drive their evolution, being an important factor for these relatively old sources. We show the consistency of this picture and remark some unsolved problems and caveats within it., Comment: 22 pp., 2 Figs, to appear in the special volume "Pulsar Astronomy with China's Facilities", Eds. R.X.Xu, Z. Gao, J.E. Horvath anbd C.A.Z. Vasconcellos (World Scientifica, Singapore, 2024)

Post-training quantization (PTQ) reduces the memory footprint of LLMs by quantizing their weights to low-precision. In this work, we introduce QuIP#, a weight-only PTQ method that achieves state-of-the-art results in extreme compression regimes ($\le$ 4 bits per weight) using three novel techniques. First, QuIP# improves QuIP's (Chee et al., 2023) incoherence processing by using the randomized Hadamard transform, which is faster and has better theoretical properties. Second, QuIP# uses vector quantization to take advantage of the ball-shaped sub-Gaussian distribution that incoherent weights possess: specifically, we introduce a set of hardware-efficient codebooks based on the highly symmetric $E_8$ lattice, which achieves the optimal 8-dimension unit ball packing. Third, QuIP# uses fine-tuning to improve fidelity to the original model. Our experiments show that QuIP# outperforms existing PTQ methods, enables new behaviors in PTQ scaling, and supports fast inference. Our code can be found at https://github.com/Cornell-RelaxML/quip-sharp., Comment: ICML 2024

In this paper, we investigate a nanomechanically induced transparency (NIT) effects that arises from the coupling of a nanoelectromechanical system and a trapped ion. By confining the ion in mesoscopic traps and capacitively coupling it with a nanoelectromechanical system suspended as electrodes, the research is intricately focussed on the implications of including the ion's degrees of freedom. The Lamb--Dicke approximation is crucial to understanding the effects of phonon exchange with electronic qubits and revealing transparency phenomena in this unique coupling. The results underline the importance of the Lamb--Dicke approximation in modelling the effects of transparency windows in nanoelectromechanical systems.

Pair production of helicity-polarized weak bosons $(V_\lambda=W^\pm_\lambda, Z_\lambda)$ from gluon fusion $(gg\to V_\lambda V'_{\lambda'})$ and weak boson fusion $(V_1V_2\to V_\lambda V'_{\lambda'})$ are powerful probes of the Standard Model, new physics, and properties of quantum systems. Measuring cross sections of polarized processes is a chief objective of the Large Hadron Collider's (LHC) Run 3 and high luminosity programs, but progress is limited by the simulation tools that are presently available. We propose a method for computing polarized cross sections that works by directly modifying Feynman rules instead of (squared) amplitudes. The method is applicable to loop-induced processes, and can capture the interference between arbitrary polarization configurations, interference with non-resonant diagrams, as well as off-shell/finite-width effects. By construction, previous results that work at the (squared) amplitude level are recoverable. As a demonstration, we report the prospect of observing and studying polarized $Z_\lambda Z_{\lambda'}$ pairs when produced via gluon fusion and electroweak processes in final-states with four charged leptons at the LHC, using the new method to simulate the gluon fusion process. Our Feynman rules are publicly available as a set of \textit{Universal FeynRules Object} libraries called \texttt{SM\_Loop\_VPolar}., Comment: Published version: additional discussion, results, and references. 16 pages, 6 figures, FeynRules UFO available from https://feynrules.irmp.ucl.ac.be/wiki/VPolarization, mg5amc configuration files available from https://gitlab.cern.ch/riruiz/public-projects/-/tree/master/VPolar_ggZZ