Search

The circumgalactic medium (CGM) in $\gtrsim 10^{12}$ $\mathrm{M}_{\odot}$ halos is dominated by a hot phase ($T \gtrsim 10^{6}$ K). While many models exist for the hot gas structure, there is as yet no consensus. We compare cooling flow models, in which the hot CGM flows inward due to radiative cooling, to the CGM of $\sim 10^{12}-10^{13}$ $\mathrm{M}_{\odot}$ halos in galaxy formation simulations from the FIRE project at $z\sim0$. The simulations include realistic cosmological evolution and feedback from stars but neglect AGN feedback. At both mass scales, CGM inflows are typically dominated by the hot phase rather than by the `precipitation' of cold gas. Despite being highly idealized, we find that cooling flows describe $\sim 10^{13}$ $\mathrm{M}_{\odot}$ halos very well, with median agreement in the density and temperature profiles of $\sim 20\%$ and $\sim 10\%$, respectively. This indicates that stellar feedback has little impact on CGM scales in those halos. For $\sim 10^{12}$ $\mathrm{M}_{\odot}$ halos, the thermodynamic profiles are also accurately reproduced in the outer CGM. For some of these lower-mass halos, cooling flows significantly overpredict the hot gas density in the inner CGM. This could be due to multidimensional angular momentum effects not well captured by our 1D cooling flow models and/or to the larger cold gas fractions in these regions. Turbulence, which contributes $\sim 10-40\%$ of the total pressure, must be included to accurately reproduce the temperature profiles. Overall, cooling flows predict entropy profiles in better agreement with the FIRE simulations than other idealized models in the literature., Comment: 20 pages, 14 figures. Submitted to MNRAS

Effective analysis of tabular data still poses a significant problem in deep learning, mainly because features in tabular datasets are often heterogeneous and have different levels of relevance. This work introduces TabSeq, a novel framework for the sequential ordering of features, addressing the vital necessity to optimize the learning process. Features are not always equally informative, and for certain deep learning models, their random arrangement can hinder the model's learning capacity. Finding the optimum sequence order for such features could improve the deep learning models' learning process. The novel feature ordering technique we provide in this work is based on clustering and incorporates both local ordering and global ordering. It is designed to be used with a multi-head attention mechanism in a denoising autoencoder network. Our framework uses clustering to align comparable features and improve data organization. Multi-head attention focuses on essential characteristics, whereas the denoising autoencoder highlights important aspects by rebuilding from distorted inputs. This method improves the capability to learn from tabular data while lowering redundancy. Our research, demonstrating improved performance through appropriate feature sequence rearrangement using raw antibody microarray and two other real-world biomedical datasets, validates the impact of feature ordering. These results demonstrate that feature ordering can be a viable approach to improved deep learning of tabular data., Comment: This paper has been accepted for presentation at the 27th International Conference on Pattern Recognition (ICPR 2024) in Kolkata, India

DNA-Encoded Libraries (DEL) are combinatorial small molecule libraries that offer an efficient way to characterize diverse chemical spaces. Selection experiments using DELs are pivotal to drug discovery efforts, enabling high-throughput screens for hit finding. However, limited availability of public DEL datasets hinders the advancement of computational techniques designed to process such data. To bridge this gap, we present KinDEL, one of the first large, publicly available DEL datasets on two kinases: Mitogen-Activated Protein Kinase 14 (MAPK14) and Discoidin Domain Receptor Tyrosine Kinase 1 (DDR1). Interest in this data modality is growing due to its ability to generate extensive supervised chemical data that densely samples around select molecular structures. Demonstrating one such application of the data, we benchmark different machine learning techniques to develop predictive models for hit identification; in particular, we highlight recent structure-based probabilistic approaches. Finally, we provide biophysical assay data, both on- and off-DNA, to validate our models on a smaller subset of molecules. Data and code for our benchmarks can be found at: https://github.com/insitro/kindel.

The vision for 6G aims to enhance network capabilities with faster data rates, near-zero latency, and higher capacity, supporting more connected devices and seamless experiences within an intelligent digital ecosystem where artificial intelligence (AI) plays a crucial role in network management and data analysis. This advancement seeks to enable immersive mixed-reality experiences, holographic communications, and smart city infrastructures. However, the expansion of 6G raises critical security and privacy concerns, such as unauthorized access and data breaches. This is due to the increased integration of IoT devices, edge computing, and AI-driven analytics. This paper provides a comprehensive overview of 6G protocols, focusing on security and privacy, identifying risks, and presenting mitigation strategies. The survey examines current risk assessment frameworks and advocates for tailored 6G solutions. We further discuss industry visions, government projects, and standardization efforts to balance technological innovation with robust security and privacy measures.

We present a practical distillation approach to fine-tune LLMs for invoking tools in real-time applications. We focus on visual editing tasks; specifically, we modify images and videos by interpreting user stylistic requests, specified in natural language ("golden hour"), using an LLM to select the appropriate tools and their parameters to achieve the desired visual effect. We found that proprietary LLMs such as GPT-3.5-Turbo show potential in this task, but their high cost and latency make them unsuitable for real-time applications. In our approach, we fine-tune a (smaller) student LLM with guidance from a (larger) teacher LLM and behavioral signals. We introduce offline metrics to evaluate student LLMs. Both online and offline experiments show that our student models manage to match the performance of our teacher model (GPT-3.5-Turbo), significantly reducing costs and latency. Lastly, we show that fine-tuning was improved by 25% in low-data regimes using augmentation., Comment: EMNLP 2024

We investigate the potential of using gravitational wave (GW) signals from rotating core-collapse supernovae to probe the equation of state (EOS) of nuclear matter. By generating GW signals from simulations with various EOSs, we train machine learning models to classify them and evaluate their performance. Our study builds on previous work by examining how different machine learning models, parameters, and data preprocessing techniques impact classification accuracy. We test convolutional and recurrent neural networks, as well as six classical algorithms: random forest, support vector machines, naive Bayes, logistic regression, k-nearest neighbors, and eXtreme gradient boosting. All models, except naive Bayes, achieve over 90 per cent accuracy on our dataset. Additionally, we assess the impact of approximating the GW signal using the general relativistic effective potential (GREP) on EOS classification. We find that models trained on GREP data exhibit low classification accuracy. However, normalizing time by the peak signal frequency, which partially compensates for the absence of the time dilation effect in GREP, leads to a notable improvement in accuracy., Comment: Submitted to Machine Learning: Science and Technology

Based upon a combined formalism of Schwinger-Dyson and Bethe-Salpeter equations in quantum chromodynamics (QCD), we propose a QCD kindred algebraic model for the dressed quark propagator, for the Bethe-Salpeter amplitude of the pion and the electromagnetic quark-photon interaction vertex. We then compute the $\gamma^{*}\pi^0\gamma$ transition form factor $G^{\gamma^{*}\pi^0\gamma}(Q^2)$ for a wide range of photon momentum transfer squared $Q^2$. The quark propagator is expanded out in its perturbative functional form but with dynamically generated dressed quark mass. It has complex conjugate pole singularities in the complex-momentum plane which is motivated by the solution of the quark gap equation with rainbow-ladder truncation of the infinite set of Schwinger-Dyson equations. This complex pole singularity structure of the quark propagator can be associated with a signal of confinement which prevents quarks to become stable asymptotic states. The Bethe-Salpeter amplitude is expressed without a spectral density function, which encapsulate its low and large momentum behaviour. The QCD evolution of the distribution amplitude is also incorporated into our model through the direct implementation of Efremov-Radyushkin-Brodsky-Lepage evolution equations. We include the effects of the quark anomalous magnetic moment in the description of the quark-photon vertex whose infrared enhancement is known to dictate hadronic properties. Once the QCD kindred model is constructed, we calculate the form factor $G^{\gamma^{*}\pi^0\gamma}(Q^2)$ and find it consistent with direct QCD-based studies as well as most available experimental data. It slightly exceeds the conformal limit for large $Q^2$ which might be attributed to the scaling violations in QCD. The associated interaction radius and neutral pion decay width turn out to be compatible with experimental data., Comment: 11 pages, 4 figures

Traditional star formation subgrid models implemented in cosmological galaxy formation simulations, such as that of Springel & Hernquist (2003, hereafter SH03), employ adjustable parameters to satisfy constraints measured in the local Universe. In recent years, however, theory and spatially-resolved simulations of the turbulent, multiphase, star-forming ISM have begun to produce new first-principles models, which when fully developed can replace traditional subgrid prescriptions. This approach has advantages of being physically motivated and predictive rather than empirically tuned, and allowing for varying environmental conditions rather than being tied to local Universe conditions. As a prototype of this new approach, by combining calibrations from the TIGRESS numerical framework with the Pressure-Regulated Feedback-Modulated (PRFM) theory, simple formulae can be obtained for both the gas depletion time and an effective equation of state. Considering galaxies in TNG50, we compare the "native" simulation outputs with post-processed predictions from PRFM. At TNG50 resolution, the total midplane pressure is nearly equal to the total ISM weight, indicating that galaxies in TNG50 are close to satisfying vertical equilibrium. The measured gas scale height is also close to theoretical equilibrium predictions. The slopes of the effective equations of states are similar, but with effective velocity dispersion normalization from SH03 slightly larger than that from current TIGRESS simulations. Because of this and the decrease in PRFM feedback yield at high pressure, the PRFM model predicts shorter gas depletion times than the SH03 model at high densities and redshift. Our results represent a first step towards implementing new, numerically calibrated subgrid algorithms in cosmological galaxy formation simulations., Comment: 33 pages, 12 figures, Accepted for publication in ApJ. This is a Learning the Universe Publication. All codes and data used to produce this work can be found at the following $\href{https://github.com/sultan-hassan/tng50-post-processing-prfm}{GitHub \,Link.}$

We present a suite of six high-resolution chemo-dynamical simulations of isolated galaxies, spanning observed disk-dominated environments on the star-forming main sequence, as well as quenched, bulge-dominated environments. We compare and contrast the physics driving star formation and stellar feedback amongst the galaxies, with a view to modeling these processes in cosmological simulations. We find that the mass-loading of galactic outflows is coupled to the clustering of supernova explosions, which varies strongly with the rate of galactic rotation $\Omega = v_c/R$ via the Toomre length, leading to smoother gas disks in the bulge-dominated galaxies. This sets an equation of state in the star-forming gas that also varies strongly with $\Omega$, so that the bulge-dominated galaxies have higher mid-plane densities, lower velocity dispersions, and higher molecular gas fractions than their main sequence counterparts. The star formation rate in five out of six galaxies is independent of $\Omega$, and is consistent with regulation by the mid-plane gas pressure alone. In the sixth galaxy, which has the most centrally-concentrated bulge and thus the highest $\Omega$, we reproduce dynamical suppression of the star formation efficiency (SFE) in agreement with observations. This produces a transition away from pressure-regulated star formation., Comment: This is a Learning the Universe Publication, accepted for publication in ApJ. 29 pages, 14 figures, and analysis code at: https://github.com/sjeffreson/pressure_regulated_SF_analysis

The $\mathcal{V}\to\mathcal{P}\gamma$ and $\eta(\eta^\prime) \to \gamma\gamma$ decays are evaluated within a Dyson-Schwinger and Bethe-Salpeter equations framework (here $\mathcal{V}=\{\rho^{\pm},K^{\star\pm},\phi\}$ and $\mathcal{P}=\{\pi^{\pm},K^{\pm},\eta,\eta^{\prime}\}$). The so-called impulse approximation (IA) is employed in the computation of the decay constants involved and decay widths, and so in the estimation of the associated charge and interaction radii. For their part, the required propagators and vertices stem from a contact interaction model, embedded within a beyond rainbow-ladder (RL) truncation that accounts for the typical ladder exchanges, quark anomalous magnetic moment, as well as the non-Abelian anomaly. While the examined transitions produce decay widths plainly compatible with the available experimental data, those processes involving the $\eta-\eta'$ mesons highlight the incompleteness of the IA when considering beyond RL effects in the interaction kernels., Comment: 9 pages

This study explores the use of historical data from Global Navigation Satellite System (GNSS) scintillation monitoring receivers to predict the severity of amplitude scintillation, a phenomenon where electron density irregularities in the ionosphere cause fluctuations in GNSS signal power. These fluctuations can be measured using the S4 index, but real-time data is not always available. The research focuses on developing a machine learning (ML) model that can forecast the intensity of amplitude scintillation, categorizing it into low, medium, or high severity levels based on various time and space-related factors. Among six different ML models tested, the XGBoost model emerged as the most effective, demonstrating a remarkable 77% prediction accuracy when trained with a balanced dataset. This work underscores the effectiveness of machine learning in enhancing the reliability and performance of GNSS signals and navigation systems by accurately predicting amplitude scintillation severity.

Aligning large language models (LLMs) with a human reasoning approach ensures that LLMs produce morally correct and human-like decisions. Ethical concerns are raised because current models are prone to generating false positives and providing malicious responses. To contribute to this issue, we have curated an ethics dataset named Dataset for Aligning Reasons (DFAR), designed to aid in aligning language models to generate human-like reasons. The dataset comprises statements with ethical-unethical labels and their corresponding reasons. In this study, we employed a unique and novel fine-tuning approach that utilizes ethics labels and their corresponding reasons (L+R), in contrast to the existing fine-tuning approach that only uses labels (L). The original pre-trained versions, the existing fine-tuned versions, and our proposed fine-tuned versions of LLMs were then evaluated on an ethical-unethical classification task and a reason-generation task. Our proposed fine-tuning strategy notably outperforms the others in both tasks, achieving significantly higher accuracy scores in the classification task and lower misalignment rates in the reason-generation task. The increase in classification accuracies and decrease in misalignment rates indicate that the L+R fine-tuned models align more with human ethics. Hence, this study illustrates that injecting reasons has substantially improved the alignment of LLMs, resulting in more human-like responses. We have made the DFAR dataset and corresponding codes publicly available at https://github.com/apurba-nsu-rnd-lab/DFAR., Comment: Accepted in ICPR 2024

Interpretable classification of time series presents significant challenges in high dimensions. Traditional feature selection methods in the frequency domain often assume sparsity in spectral density matrices (SDMs) or their inverses, which can be restrictive for real-world applications. In this article, we propose a model-based approach for classifying high-dimensional stationary time series by assuming sparsity in the difference between inverse SDMs. Our approach emphasizes the interpretability of model parameters, making it especially suitable for fields like neuroscience, where understanding differences in brain network connectivity across various states is crucial. The estimators for model parameters demonstrate consistency under appropriate conditions. We further propose using standard deep learning optimizers for parameter estimation, employing techniques such as mini-batching and learning rate scheduling. Additionally, we introduce a method to screen the most discriminatory frequencies for classification, which exhibits the sure screening property under general conditions. The flexibility of the proposed model allows the significance of covariates to vary across frequencies, enabling nuanced inferences and deeper insights into the underlying problem. The novelty of our method lies in the interpretability of the model parameters, addressing critical needs in neuroscience. The proposed approaches have been evaluated on simulated examples and the `Alert-vs-Drowsy' EEG dataset.

Iris recognition is widely used in several fields such as mobile phones, financial transactions, identification cards, airport security, international border control, voter registration for living persons. However, the possibility of identifying deceased individuals based on their iris patterns has emerged recently as a supplementary or alternative method valuable in forensic analysis. Simultaneously, it poses numerous new technological challenges and one of the most challenging among them is the image segmentation stage as conventional iris recognition approaches have struggled to reliably execute it. This paper presents and compares Deep Learning (DL) models designed for segmenting iris images collected from the deceased subjects, by training SegNet and DeepLabV3+ semantic segmentation methods where using VGG19, ResNet18, ResNet50, MobileNetv2, Xception, or InceptionResNetv2 as backbones. In this study, our experiments demonstrate that our proposed method effectively learns and identifies specific deformations inherent in post-mortem samples and providing a significant improvement in accuracy. By employing our novel method MobileNetv2 as the backbone of DeepLabV3+ and replacing the final layer with a hybrid loss function combining Boundary and Dice loss, we achieve Mean Intersection over Union of 95.54% on the Warsaw-BioBase-PostMortem-Iris-v1 dataset. To the best of our knowledge, this study provides the most extensive evaluation of DL models for post-mortem iris segmentation., Comment: submitted to ijcb 2024 special session

Ear recognition as a biometric modality is becoming increasingly popular, with promising broader application areas. While current applications involve adults, one of the challenges in ear recognition for children is the rapid structural changes in the ear as they age. This work introduces a foundational longitudinal dataset collected from children aged 4 to 14 years over a 2.5-year period and evaluates ear recognition performance in this demographic. We present a deep learning based approach for ear recognition, using an ensemble of VGG16 and MobileNet, focusing on both adult and child datasets, with an emphasis on longitudinal evaluation for children., Comment: Submitted to Biosig 2024

First time in six decades, uncountable infinite exact solutions of FitzHugh-Nagumo model with diffusion have been found. FitzHugh-Nagumo model is a nonlinear dynamical system applicable to neurosciences, chemical kinetics, cell division, population dynamics, electronics, epidemiology, cardiac physiology and pattern formation. Non-classical symmetry analysis has been carried out to find invariant solutions. In many ways this finding makes the corpus of asymptotics and numerics obsolete. Non singular, physically stable and meaningful solutions could now be found without distorting the actual model or introducing forced conditions.

We present ALLaM: Arabic Large Language Model, a series of large language models to support the ecosystem of Arabic Language Technologies (ALT). ALLaM is carefully trained considering the values of language alignment and knowledge transfer at scale. Our autoregressive decoder-only architecture models demonstrate how second-language acquisition via vocabulary expansion and pretraining on a mixture of Arabic and English text can steer a model towards a new language (Arabic) without any catastrophic forgetting in the original language (English). Furthermore, we highlight the effectiveness of using parallel/translated data to aid the process of knowledge alignment between languages. Finally, we show that extensive alignment with human preferences can significantly enhance the performance of a language model compared to models of a larger scale with lower quality alignment. ALLaM achieves state-of-the-art performance in various Arabic benchmarks, including MMLU Arabic, ACVA, and Arabic Exams. Our aligned models improve both in Arabic and English from their base aligned models.

Extractive question answering over clinical text is a crucial need to help deal with the deluge of clinical text generated in hospitals. While encoder models (e.g., BERT) have been popular for this reading comprehension task, recently encoder-decoder models (e.g., T5) are on the rise. There is also the emergence of preference optimization techniques to align decoder-only LLMs with human preferences. In this paper, we combine encoder-decoder models with the direct preference optimization (DPO) method to improve over prior state of the art for the RadQA radiology question answering task by 12-15 F1 points. To the best of our knowledge, this effort is the first to show that DPO method also works for reading comprehension via novel heuristics to generate preference data without human inputs.

Given the unique role played by the gravitational form factors (GFFs) in unraveling the internal mechanics of hadrons, we examine the GFFs of ground state pseudoscalar mesons $\pi$, $\eta_c$, $\eta_b$ and the hypothetical {\em strangeonium} $\eta_s(s\bar{s})$. We adopt the coupled framework of Dyson-Schwinger and Bethe-Salpeter equations within a contact interaction, and employ a novel approach to the dressed amputated meson-meson scattering amplitude which makes connection with the energy-momentum tensor and with the GFFs. The resulting GFFs fulfill the anticipated symmetry constraints. The corresponding charge and mass radii and the $D-$term are also computed. We show that the $D-$term for the pseudoscalar mesons is bounded within the $(-1, -1/3)$ range; these bounds correspond to the massless (chiral limit) and infinitely massive cases, respectively. Considering the current interest in the GFFs, understanding the \textit{D}-term of pseudoscalar mesons and their GFFs can provide an important first step for future endeavors in the field., Comment: 10 pages, 5 figures

The expanding financial markets of the Arab world require sophisticated Arabic NLP tools. To address this need within the banking domain, the Arabic Financial NLP (AraFinNLP) shared task proposes two subtasks: (i) Multi-dialect Intent Detection and (ii) Cross-dialect Translation and Intent Preservation. This shared task uses the updated ArBanking77 dataset, which includes about 39k parallel queries in MSA and four dialects. Each query is labeled with one or more of a common 77 intents in the banking domain. These resources aim to foster the development of robust financial Arabic NLP, particularly in the areas of machine translation and banking chat-bots. A total of 45 unique teams registered for this shared task, with 11 of them actively participated in the test phase. Specifically, 11 teams participated in Subtask 1, while only 1 team participated in Subtask 2. The winning team of Subtask 1 achieved F1 score of 0.8773, and the only team submitted in Subtask 2 achieved a 1.667 BLEU score.

In this paper, we show different fine-tuning methods for Stable Diffusion XL; this includes inference steps, and caption customization for each image to align with generating images in the style of a commercial 2D icon training set. We also show how important it is to properly define what "high-quality" really is especially for a commercial-use environment. As generative AI models continue to gain widespread acceptance and usage, there emerge many different ways to optimize and evaluate them for various applications. Specifically text-to-image models, such as Stable Diffusion XL and DALL-E 3 require distinct evaluation practices to effectively generate high-quality icons according to a specific style. Although some images that are generated based on a certain style may have a lower FID score (better), we show how this is not absolute in and of itself even for rasterized icons. While FID scores reflect the similarity of generated images to the overall training set, CLIP scores measure the alignment between generated images and their textual descriptions. We show how FID scores miss significant aspects, such as the minority of pixel differences that matter most in an icon, while CLIP scores result in misjudging the quality of icons. The CLIP model's understanding of "similarity" is shaped by its own training data; which does not account for feature variation in our style of choice. Our findings highlight the need for specialized evaluation metrics and fine-tuning approaches when generating high-quality commercial icons, potentially leading to more effective and tailored applications of text-to-image models in professional design contexts., Comment: 11 pages, 22 figures

The accurate evaluation of left atrial fibrosis via high-quality 3D Late Gadolinium Enhancement (LGE) MRI is crucial for atrial fibrillation management but is hindered by factors like patient movement and imaging variability. The pursuit of automated LGE MRI quality assessment is critical for enhancing diagnostic accuracy, standardizing evaluations, and improving patient outcomes. The deep learning models aimed at automating this process face significant challenges due to the scarcity of expert annotations, high computational costs, and the need to capture subtle diagnostic details in highly variable images. This study introduces HAMIL-QA, a multiple instance learning (MIL) framework, designed to overcome these obstacles. HAMIL-QA employs a hierarchical bag and sub-bag structure that allows for targeted analysis within sub-bags and aggregates insights at the volume level. This hierarchical MIL approach reduces reliance on extensive annotations, lessens computational load, and ensures clinically relevant quality predictions by focusing on diagnostically critical image features. Our experiments show that HAMIL-QA surpasses existing MIL methods and traditional supervised approaches in accuracy, AUROC, and F1-Score on an LGE MRI scan dataset, demonstrating its potential as a scalable solution for LGE MRI quality assessment automation. The code is available at: $\href{https://github.com/arf111/HAMIL-QA}{\text{this https URL}}$, Comment: Accepted to MICCAI2024, 10 pages, 2 figures

We find some exact solutions for constant-density and quark matter equations of state in stellar structure models framed within the $f(R,T)=R+\lambda \kappa^2 T$ theory of gravity, where $R$ is the curvature scalar, $T$ the trace of the stress-energy tensor, and $\lambda$ some constant. These solutions correspond to specific values of the constant $\lambda$, and represent different compactness states of the corresponding stars, though only those made of quark matter can be regarded as physical. The latter modify the compactness (Buchdahl) limit of neutron stars upwards for $\lambda>0$ until saturating the one of black holes. Our results show that it is possible to find useful insights on stellar structure in this class of theories, a fact that could be used for obtaining constraints on limiting masses such as the minimum hydrogen burning mass., Comment: 6 pages, 1 figure, revtex style

The aim of this contribution is to address the convergence study of a time and space approximation scheme for an Allen-Cahn problem with constraint and perturbed by a multiplicative noise of It\^o type. The problem is set in a bounded domain of $\mathbb{R}^d$ (with $d=2$ or $3$) and homogeneous Neumann boundary conditions are considered. The employed strategy consists in building a numerical scheme on a regularized version \`a la Moreau-Yosida of the constrained problem, and passing to the limit simultaneously with respect to the regularization parameter and the time and space steps, denoted respectively by $\epsilon$, $\Delta t$ and $h$. Combining a semi-implicit Euler-Maruyama time discretization with a Two-Point Flux Approximation (TPFA) scheme for the spatial variable, one is able to prove, under the assumption $\Delta t=\mathcal{O}(\epsilon^{2+\theta})$ for a positive $\theta$, the convergence of such a $(\epsilon, \Delta t, h)$ scheme towards the unique weak solution of the initial problem, \textit{ a priori} strongly in $L^2(\Omega;L^2(0,T;L^2(\Lambda)))$ and \textit{a posteriori} also strongly in $L^{p}(0,T; L^2(\Omega\times \Lambda))$ for any finite $p\geq 1$., Comment: arXiv admin note: text overlap with arXiv:2304.02259

Large Language Models (LLMs) have significantly advanced the field of information retrieval, particularly for reranking. Listwise LLM rerankers have showcased superior performance and generalizability compared to existing supervised approaches. However, conventional listwise LLM reranking methods lack efficiency as they provide ranking output in the form of a generated ordered sequence of candidate passage identifiers. Further, they are trained with the typical language modeling objective, which treats all ranking errors uniformly--potentially at the cost of misranking highly relevant passages. Addressing these limitations, we introduce FIRST, a novel listwise LLM reranking approach leveraging the output logits of the first generated identifier to directly obtain a ranked ordering of the candidates. Further, we incorporate a learning-to-rank loss during training, prioritizing ranking accuracy for the more relevant passages. Empirical results demonstrate that FIRST accelerates inference by 50% while maintaining a robust ranking performance with gains across the BEIR benchmark. Finally, to illustrate the practical effectiveness of listwise LLM rerankers, we investigate their application in providing relevance feedback for retrievers during inference. Our results show that LLM rerankers can provide a stronger distillation signal compared to cross-encoders, yielding substantial improvements in retriever recall after relevance feedback., Comment: Preprint

Integrated optomechanical systems are one of the leading platforms for manipulating, sensing, and distributing quantum information. The temperature increase due to residual optical absorption sets the ultimate limit on performance for these applications. In this work, we demonstrate a two-dimensional optomechanical crystal geometry, named \textbf{b-dagger}, that alleviates this problem through increased thermal anchoring to the surrounding material. Our mechanical mode operates at 7.4 GHz, well within the operation range of standard cryogenic microwave hardware and piezoelectric transducers. The enhanced thermalization combined with the large optomechanical coupling rates, $g_0/2\pi \approx 880~\mathrm{kHz}$, and high optical quality factors, $Q_\text{opt} = 2.4 \times 10^5$, enables the ground-state cooling of the acoustic mode to phononic occupancies as low as $n_\text{m} = 0.35$ from an initial temperature of 3 kelvin, as well as entering the optomechanical strong-coupling regime. Finally, we perform pulsed sideband asymmetry of our devices at a temperature below 10 millikelvin and demonstrate ground-state operation ($n_\text{m} < 0.45$) for repetition rates as high as 3 MHz. Our results extend the boundaries of optomechanical system capabilities and establish a robust foundation for the next generation of microwave-to-optical transducers with entanglement rates overcoming the decoherence rates of state-of-the-art superconducting qubits., Comment: 13 pages, 4 main figures

The evolution of the entanglement measure during Compton scattering is studied. Our analytical results show that the corresponding measure coincides with the concurrence of the two-qubit state arising after scattering. The state never collapses to a separable one, contrary to what was previously assumed. The behavior of quantum entanglement during scattering is identical to the behavior of initially classically correlated photons up to a constant factor equal to two. This is consistent with local quantum field theory, and "spooky action at a distance" is not required to explain the change in state of nonlocally entangled qubits during the measurement of one of them. Our dedicated experiment with annihilation photons confirms these results and explains the "Puzzle of Decoherence" observed recently., Comment: 11 pages, 8 figures

We develop a method for training small-scale (under 100M parameter) neural information retrieval models with as few as 10 gold relevance labels. The method depends on generating synthetic queries for documents using a language model (LM), and the key step is that we automatically optimize the LM prompt that is used to generate these queries based on training quality. In experiments with the BIRCO benchmark, we find that models trained with our method outperform RankZephyr and are competitive with RankLLama, both of which are 7B parameter models trained on over 100K labels. These findings point to the power of automatic prompt optimization for synthetic dataset generation.

In order to ensure traffic safety through a reduction in fatalities and accidents, vehicle speed detection is essential. Relentless driving practices are discouraged by the enforcement of speed restrictions, which are made possible by accurate monitoring of vehicle speeds. Road accidents remain one of the leading causes of death in Bangladesh. The Bangladesh Passenger Welfare Association stated in 2023 that 7,902 individuals lost their lives in traffic accidents during the course of the year. Efficient vehicle speed detection is essential to maintaining traffic safety. Reliable speed detection can also help gather important traffic data, which makes it easier to optimize traffic flow and provide safer road infrastructure. The YOLOv8 model can recognize and track cars in videos with greater speed and accuracy when trained under close supervision. By providing insights into the application of supervised learning in object identification for vehicle speed estimation and concentrating on the particular traffic conditions and safety concerns in Bangladesh, this work represents a noteworthy contribution to the area. The MAE was 3.5 and RMSE was 4.22 between the predicted speed of our model and the actual speed or the ground truth measured by the speedometer Promising increased efficiency and wider applicability in a variety of traffic conditions, the suggested solution offers a financially viable substitute for conventional approaches.

In the construction sector, ensuring worker safety is of the utmost significance. In this study, a deep learning-based technique is presented for identifying safety gear worn by construction workers, such as helmets, goggles, jackets, gloves, and footwears. The recommended approach uses the YOLO v7 (You Only Look Once) object detection algorithm to precisely locate these safety items. The dataset utilized in this work consists of labeled images split into training, testing and validation sets. Each image has bounding box labels that indicate where the safety equipment is located within the image. The model is trained to identify and categorize the safety equipment based on the labeled dataset through an iterative training approach. We used custom dataset to train this model. Our trained model performed admirably well, with good precision, recall, and F1-score for safety equipment recognition. Also, the model's evaluation produced encouraging results, with a mAP@0.5 score of 87.7\%. The model performs effectively, making it possible to quickly identify safety equipment violations on building sites. A thorough evaluation of the outcomes reveals the model's advantages and points up potential areas for development. By offering an automatic and trustworthy method for safety equipment detection, this research makes a contribution to the fields of computer vision and workplace safety. The proposed deep learning-based approach will increase safety compliance and reduce the risk of accidents in the construction industry

In this study we investigate novel hybrid structure of silicon nanowires (SiNWs) coated with PANI:metaloxide(MO$_x$) nanoparticles i.e., WO$_3$ and TiO$_2$. The SiNWs were fabricated using MACE, whereas PANI:MO$_x$ were deposited using chemical oxidative polymerization method on SiNWs. To this date little attempts has been done to utilize such hybrid structures for respiratory sensing. The structures were characterized using RAMAN spectroscopy, X-ray diffraction, Electron disperssive spectroscopy, and Scanning electron microscopy. The electrical characterization to obtain respiratory sensing reveals excellent response compared to those obtained for SiNWs:MO$_x$ and SiNWs:PANI. Such enhancement in sensitivity is attributed to formation p-n heterojunction along side with wider conduction channel provided of PANI, increased porosity in SiNWs/PANI:WO$_3$ hybrid structures, providing active sites, increased oxygen vacancies and large surface area compared to that of pure MO$_x$ nanoparticles. Further, an improved drift in base line and sensor stability was established for the structure with PANI:WO$_3$ as compared to the PANI:TiO$_2$., Comment: 12 figures, 16 pages, 44 references

We present a new simulation suite for the star-forming interstellar medium (ISM) in galactic disks using the TIGRESS-NCR framework. Distinctive aspects of our simulation suite are: (1) sophisticated and comprehensive numerical treatments of essential physical processes including magnetohydrodynamics, self-gravity, and galactic differential rotation, as well as photochemistry, cooling, and heating coupled with ray-tracing UV radiation transfer and resolved supernova feedback and (2) wide parameter coverage including metallicity over $Z'\equiv Z/Z_\odot\sim0.1-3$, gas surface density $\Sigma_{\rm gas}\sim5-150 M_{\odot}{\rm pc^{-2}}$, and stellar surface density $\Sigma_{\rm star}\sim 1-50 M_{\odot}{\rm pc^{-2}}$. The range of emergent star formation rate surface density is $\Sigma_{\rm SFR}\sim 10^{-4}-0.5 M_{\odot}{\rm kpc^{-2}yr^{-1}}$ and ISM total midplane pressure is $P_{\rm tot}/k_B=10^3-10^6{\rm cm^{-3}K}$, with $P_{\rm tot}$ equal to the ISM weight $W$. For given $\Sigma_{\rm gas}$ and $\Sigma_{\rm star}$, we find $\Sigma_{\rm SFR} \propto Z'^{0.3}$. We provide an interpretation based on the pressure-regulated feedback-modulated (PRFM) star formation theory. We characterize feedback modulation in terms of the yield $\Upsilon$, defined as the ratio of each stress to $\Sigma_{\rm SFR}$. The thermal feedback yield varies sensitively with both weight and metallicity as $\Upsilon_{\rm th}\propto W^{-0.46}Z'^{-0.53}$, while the combined turbulent and magnetic feedback yield shows weaker dependence $\Upsilon_{\rm turb+mag}\propto W^{-0.22}Z'^{-0.18}$. The reduction in $\Sigma_{\rm SFR}$ at low metallicity is due mainly to enhanced thermal feedback yield, resulting from reduced attenuation of UV radiation. With the metallicity-dependent calibrations we provide, PRFM theory can be used for a new subgrid star formation prescription in cosmological simulations where the ISM is unresolved., Comment: Resubmitted to ApJ after minor revision

The world's healthcare systems and mental health agencies face both a growing demand for youth mental health services, alongside a simultaneous challenge of limited resources. Here, we focus on frontline crisis support, where Crisis Responders (CRs) engage in conversations for youth mental health support and assign an issue tag to each conversation. In this study, we develop FAIIR (Frontline Assistant: Issue Identification and Recommendation), an advanced tool leveraging an ensemble of domain-adapted and fine-tuned transformer models trained on a large conversational dataset comprising 780,000 conversations. The primary aim is to reduce the cognitive burden on CRs, enhance the accuracy of issue identification, and streamline post-conversation administrative tasks. We evaluate FAIIR on both retrospective and prospective conversations, emphasizing human-in-the-loop design with active CR engagement for model refinement, consensus-building, and overall assessment. Our results indicate that FAIIR achieves an average AUCROC of 94%, a sample average F1-score of 64%, and a sample average recall score of 81% on the retrospective test set. We also demonstrate the robustness and generalizability of the FAIIR tool during the silent testing phase, with less than a 2% drop in all performance metrics. Notably, CRs' responses exhibited an overall agreement of 90.9% with FAIIR's predictions. Furthermore, expert agreement with FAIIR surpassed their agreement with the original labels. To conclude, our findings indicate that assisting with the identification of issues of relevance helps reduce the burden on CRs, ensuring that appropriate resources can be provided and that active rescues and mandatory reporting can take place in critical situations requiring immediate de-escalation.

In recent years, convolutional neural networks (CNNs) have achieved remarkable advancement in the field of remote sensing image super-resolution due to the complexity and variability of textures and structures in remote sensing images (RSIs), which often repeat in the same images but differ across others. Current deep learning-based super-resolution models focus less on high-frequency features, which leads to suboptimal performance in capturing contours, textures, and spatial information. State-of-the-art CNN-based methods now focus on the feature extraction of RSIs using attention mechanisms. However, these methods are still incapable of effectively identifying and utilizing key content attention signals in RSIs. To solve this problem, we proposed an advanced feature extraction module called Channel and Spatial Attention Feature Extraction (CSA-FE) for effectively extracting the features by using the channel and spatial attention incorporated with the standard vision transformer (ViT). The proposed method trained over the UCMerced dataset on scales 2, 3, and 4. The experimental results show that our proposed method helps the model focus on the specific channels and spatial locations containing high-frequency information so that the model can focus on relevant features and suppress irrelevant ones, which enhances the quality of super-resolved images. Our model achieved superior performance compared to various existing models., Comment: Preprint of paper from The 21st International Conference on Electrical Engineering/Electronics, Computer, Telecommunications and Information Technology or ECTI-CON 2024, Khon Kaen, Thailand

The advent of Large Language Models (LLMs) has revolutionized various domains of artificial intelligence, including the realm of software engineering. In this research, we evaluate the efficacy of pre-trained LLMs in replicating the tasks traditionally performed by developers in response to code review comments. We provide code contexts to five popular LLMs and obtain the suggested code-changes (patch sets) derived from real-world code-review comments. The performance of each model is meticulously assessed by comparing their generated patch sets against the historical data of human-generated patch-sets from the same repositories. This comparative analysis aims to determine the accuracy, relevance, and depth of the LLMs' feedback, thereby evaluating their readiness to support developers in responding to code-review comments. Novelty: This particular research area is still immature requiring a substantial amount of studies yet to be done. No prior research has compared the performance of existing Large Language Models (LLMs) in code-review comments. This in-progress study assesses current LLMs in code review and paves the way for future advancements in automated code quality assurance, reducing context-switching overhead due to interruptions from code change requests., Comment: 2 pages