Search

Transformers have revolutionized Computer Vision (CV) and Natural Language Processing (NLP) through self-attention mechanisms. However, due to their complexity, their latent token representations are often difficult to interpret. We introduce a novel framework that interprets Transformer embeddings, uncovering meaningful semantic patterns within them. Based on this framework, we demonstrate that zero-shot unsupervised semantic segmentation can be performed effectively without any fine-tuning using a model pre-trained for tasks other than segmentation. Our method reveals the inherent capacity of Transformer models for understanding input semantics and achieves state-of-the-art performance in semantic segmentation, outperforming traditional segmentation models. Specifically, our approach achieves an accuracy of 67.2 % and an mIoU of 32.9 % on the COCO-Stuff dataset, as well as an mIoU of 51.9 % on the PASCAL VOC dataset. Additionally, we validate our interpretability framework on LLMs for text summarization, demonstrating its broad applicability and robustness.

Time series data analysis is prevalent across various domains, including finance, healthcare, and environmental monitoring. Traditional time series clustering methods often struggle to capture the complex temporal dependencies inherent in such data. In this paper, we propose the Variational Mixture Graph Autoencoder (VMGAE), a graph-based approach for time series clustering that leverages the structural advantages of graphs to capture enriched data relationships and produces Gaussian mixture embeddings for improved separability. Comparisons with baseline methods are included with experimental results, demonstrating that our method significantly outperforms state-of-the-art time-series clustering techniques. We further validate our method on real-world financial data, highlighting its practical applications in finance. By uncovering community structures in stock markets, our method provides deeper insights into stock relationships, benefiting market prediction, portfolio optimization, and risk management., Comment: First two listed authors have equal contribution. Author ordering is determined by coin flip

Diffusion Generative Models (DGM) have rapidly surfaced as emerging topics in the field of computer vision, garnering significant interest across a wide array of deep learning applications. Despite their high computational demand, these models are extensively utilized for their superior sample quality and robust mode coverage. While research in diffusion generative models is advancing, exploration within the domain of computational pathology and its large-scale datasets has been comparatively gradual. Bridging the gap between the high-quality generation capabilities of Diffusion Generative Models and the intricate nature of pathology data, this paper presents an in-depth comparative analysis of diffusion methods applied to a pathology dataset. Our analysis extends to datasets with varying Fields of View (FOV), revealing that DGMs are highly effective in producing high-quality synthetic data. An ablative study is also conducted, followed by a detailed discussion on the impact of various methods on the synthesized histopathology images. One striking observation from our experiments is how the adjustment of image size during data generation can simulate varying fields of view. These findings underscore the potential of DGMs to enhance the quality and diversity of synthetic pathology data, especially when used with real data, ultimately increasing accuracy of deep learning models in histopathology. Code is available from https://github.com/AtlasAnalyticsLab/Diffusion4Path, Comment: Submitted paper under review

Unmanned aerial vehicles (UAVs) have the potential for time-sensitive applications. Due to wireless channel variation, received data may have an expiration time, particularly in critical situations such as rescue operations, natural disasters, or the military. Age of Information (AoI) is a metric that measures the freshness of received packets to specify the validity period of information. In addition, it is necessary to guarantee the privacy of confidential information transmission through air-to-ground links against eavesdroppers. This paper investigates UAV-assisted covert communication to minimize AoI in the presence of an aerial eavesdropper for the first time. However, to ensure the eavesdropper's error detection rate, UAV-enabled beamforming employs the power-domain non-orthogonal multiple access (PD-NOMA) technique to cover the covert user by a public user. PD-NOMA technique significantly improves the user's AoI, too. The joint optimization problem contains non-convex constraints and coupled optimization variables, including UAV trajectory, beamforming design, and the user's AoI which is challenging to derive a direct solution. We have developed an efficient alternating optimization technique to address the formulated optimization problem. Numerical results demonstrate the impact of the main parameters on the performance of the proposed communication system.

This study presents the modeling of the gravitational wave (GW) bias parameter by bridging a connection between simulated GW sources and galaxies in low redshift galaxy surveys 2MPZ and WISExSCOS (WISC). We study this connection by creating a mock GW catalog, populating galaxy surveys with binary black holes (BBHs) for different scenarios of the GW occupation fraction (or selection function) as a function of the galaxy stellar mass. We probe the observable consequences of this connection by exploring the spatial clustering of the GW sources in terms of the GW bias parameter. We consider a phenomenological broken power law model for the selection function, with a potential turnover $M_{K}$ at high stellar mass ($10^{11}$ $M_{\odot}$ in the fiducial model) where the star formation efficiency begins to drop. We vary the parameters of the selection function and find that generically the GW bias increases as $M_{K}$ increases (and gets suppressed as $M_{K}$ decreases). The change in the GW bias parameter shows a maximum change of about $30\%$ for different scenarios explored in this work in comparison to the galaxy bias. Future measurements of the GW bias can help constrain $M_{K}$ and the slopes of the selection function and thus offer insights into the underlying astrophysical processes., Comment: 29 pages (+15 pages in appendices), 15 figures (+9 figures in appendices), to be submitted to JCAP

We analyze the static potentials induced by vacuum domains for various representations in SU($4$) Yang-Mills theory within the framework of the domain structure model. By studying the interactions within the vacuum domains, we can uncover fundamental properties of the static potentials. It appears that attractions within the vacuum domains strongly adhere to Casimir scaling at intermediate distances. Conversely, the repulsions within the vacuum domains may decompose them into center vortices with the lowest magnitude of center vortex fluxes, thereby exhibiting $N$-ality at asymptotic distances., Comment: 12 pages, 12 figures

There is a lack of benchmarks for evaluating large language models (LLMs) in long-form medical question answering (QA). Most existing medical QA evaluation benchmarks focus on automatic metrics and multiple-choice questions. While valuable, these benchmarks fail to fully capture or assess the complexities of real-world clinical applications where LLMs are being deployed. Furthermore, existing studies on evaluating long-form answer generation in medical QA are primarily closed-source, lacking access to human medical expert annotations, which makes it difficult to reproduce results and enhance existing baselines. In this work, we introduce a new publicly available benchmark featuring real-world consumer medical questions with long-form answer evaluations annotated by medical doctors. We performed pairwise comparisons of responses from various open and closed-source medical and general-purpose LLMs based on criteria such as correctness, helpfulness, harmfulness, and bias. Additionally, we performed a comprehensive LLM-as-a-judge analysis to study the alignment between human judgments and LLMs. Our preliminary results highlight the strong potential of open LLMs in medical QA compared to leading closed models. Code & Data: https://github.com/lavita-ai/medical-eval-sphere, Comment: AIM-FM: Advancements in Medical Foundation Models Workshop, 38th Conference on Neural Information Processing Systems (NeurIPS 2024)

An important problem on social information sites is the recovery of ground truth from individual reports when the experts are in the minority. The wisdom of the crowd, i.e. the collective opinion of a group of individuals fails in such a scenario. However, the surprisingly popular (SP) algorithm~\cite{prelec2017solution} can recover the ground truth even when the experts are in the minority, by asking the individuals to report additional prediction reports--their beliefs about the reports of others. Several recent works have extended the surprisingly popular algorithm to an equivalent voting rule (SP-voting) to recover the ground truth ranking over a set of $m$ alternatives. However, we are yet to fully understand when SP-voting can recover the ground truth ranking, and if so, how many samples (votes and predictions) it needs. We answer this question by proposing two rank-order models and analyzing the sample complexity of SP-voting under these models. In particular, we propose concentric mixtures of Mallows and Plackett-Luce models with $G (\ge 2)$ groups. Our models generalize previously proposed concentric mixtures of Mallows models with $2$ groups, and we highlight the importance of $G > 2$ groups by identifying three distinct groups (expert, intermediate, and non-expert) from existing datasets. Next, we provide conditions on the parameters of the underlying models so that SP-voting can recover ground-truth rankings with high probability, and also derive sample complexities under the same. We complement the theoretical results by evaluating SP-voting on simulated and real datasets.

We focus on a sample of 42 sources in the vicinity of the bow-shock source IRS 1W (N-sources), located at the distance of $6.05''$ north-east of the supermassive black hole (SMBH) Sagittarius A* (Sgr A*), within the radius of $1.35''$. We present the first proper motion measurements of N-sources and find that a larger subset of N-sources (28 sources) exhibit a north-westward flying angle. These sources can be bound by an intermediate mass black hole (IMBH) or the concentration that we observe is due to a disk-like distribution projection along the line of sight. We detect the N-sources in $H$, $K_s$, and $L$' bands. The north-westward flying sources could be a bound collection of stars. We discuss a tentative existence of an IMBH or an inclined disk distribution to explain a significant overdensity of stars. The first scenario of having an IMBH implies the lower limit of $\sim 10^4~M_\odot$ for the putative IMBH. Our measurements for the first time reveal that the dense association of stars containing IRS 1W is a co-moving group of massive, young stars. This stellar association might be the remnant core of a massive stellar cluster that is currently being tidally stripped as it inspirals towards Sgr A*. The second scenario suggests that the appearance of the N-sources might be influenced by the projection of a disk-like distribution of younger He-stars and/or dust-enshrouded stars., Comment: 21 pages, 17 figures; published in the Astrophysical Journal

Colorectal cancer (CRC) is one of the few cancers that have an established dysplasia-carcinoma sequence that benefits from screening. Everyone over 50 years of age in Canada is eligible for CRC screening. About 20\% of those people will undergo a biopsy for a pre-neoplastic polyp and, in many cases, multiple polyps. As such, these polyp biopsies make up the bulk of a pathologist's workload. Developing an efficient computational model to help screen these polyp biopsies can improve the pathologist's workflow and help guide their attention to critical areas on the slide. DL models face significant challenges in computational pathology (CPath) because of the gigapixel image size of whole-slide images and the scarcity of detailed annotated datasets. It is, therefore, crucial to leverage self-supervised learning (SSL) methods to alleviate the burden and cost of data annotation. However, current research lacks methods to apply SSL frameworks to analyze pathology data effectively. This paper aims to propose an optimized Barlow Twins framework for colorectal polyps screening. We adapt its hyperparameters, augmentation strategy and encoder to the specificity of the pathology data to enhance performance. Additionally, we investigate the best Field of View (FoV) for colorectal polyps screening and propose a new benchmark dataset for CRC screening, made of four types of colorectal polyps and normal tissue, by performing downstream tasking on MHIST and NCT-CRC-7K datasets. Furthermore, we show that the SSL representations are more meaningful and qualitative than the supervised ones and that Barlow Twins benefits from the Swin Transformer when applied to pathology data. Codes are avaialble from https://github.com/AtlasAnalyticsLab/PathBT., Comment: Submission Under Review

The Generalist Repository Ecosystem Initiative (GREI), funded by the NIH, developed an AI taxonomy tailored to data repository roles to guide AI integration across repository management. It categorizes the roles into stages, including acquisition, validation, organization, enhancement, analysis, sharing, and user support, providing a structured framework for implementing AI in repository workflows.

Retrieval augmented generation (RAG) models, which integrate large-scale pre-trained generative models with external retrieval mechanisms, have shown significant success in various natural language processing (NLP) tasks. However, applying RAG models in Persian language as a low-resource language, poses distinct challenges. These challenges primarily involve the preprocessing, embedding, retrieval, prompt construction, language modeling, and response evaluation of the system. In this paper, we address the challenges towards implementing a real-world RAG system for Persian language called PersianRAG. We propose novel solutions to overcome these obstacles and evaluate our approach using several Persian benchmark datasets. Our experimental results demonstrate the capability of the PersianRAG framework to enhance question answering task in Persian.

Osteoporosis is a common condition that increases fracture risk, especially in older adults. Early diagnosis is vital for preventing fractures, reducing treatment costs, and preserving mobility. However, healthcare providers face challenges like limited labeled data and difficulties in processing medical images. This study presents a novel multi-modal learning framework that integrates clinical and imaging data to improve diagnostic accuracy and model interpretability. The model utilizes three pre-trained networks-VGG19, InceptionV3, and ResNet50-to extract deep features from X-ray images. These features are transformed using PCA to reduce dimensionality and focus on the most relevant components. A clustering-based selection process identifies the most representative components, which are then combined with preprocessed clinical data and processed through a fully connected network (FCN) for final classification. A feature importance plot highlights key variables, showing that Medical History, BMI, and Height were the main contributors, emphasizing the significance of patient-specific data. While imaging features were valuable, they had lower importance, indicating that clinical data are crucial for accurate predictions. This framework promotes precise and interpretable predictions, enhancing transparency and building trust in AI-driven diagnoses for clinical integration.

In smart manufacturing, scheduling autonomous internal logistic vehicles is crucial for optimizing operational efficiency. This paper proposes a multi-agent deep Q-network (MADQN) with a layer-based communication channel (LBCC) to address this challenge. The main goals are to minimize total job tardiness, reduce the number of tardy jobs, and lower vehicle energy consumption. The method is evaluated against nine well-known scheduling heuristics, demonstrating its effectiveness in handling dynamic job shop behaviors like job arrivals and workstation unavailabilities. The approach also proves scalable, maintaining performance across different layouts and larger problem instances, highlighting the robustness and adaptability of MADQN with LBCC in smart manufacturing., Comment: Accepted for the 5th IFAC/INSTICC INTERNATIONAL CONFERENCE ON INNOVATIVE INTELLIGENT INDUSTRIAL PRODUCTION AND LOGISTICS

Automated Facial Expression Recognition (FER) is challenging due to intra-class variations and inter-class similarities. FER can be especially difficult when facial expressions reflect a mixture of various emotions (aka compound expressions). Existing FER datasets, such as AffectNet, provide discrete emotion labels (hard-labels), where a single category of emotion is assigned to an expression. To alleviate inter- and intra-class challenges, as well as provide a better facial expression descriptor, we propose a new approach to create FER datasets through a labeling method in which an image is labeled with more than one emotion (called soft-labels), each with different confidences. Specifically, we introduce the notion of soft-labels for facial expression datasets, a new approach to affective computing for more realistic recognition of facial expressions. To achieve this goal, we propose a novel methodology to accurately calculate soft-labels: a vector representing the extent to which multiple categories of emotion are simultaneously present within a single facial expression. Finding smoother decision boundaries, enabling multi-labeling, and mitigating bias and imbalanced data are some of the advantages of our proposed method. Building upon AffectNet, we introduce AffectNet+, the next-generation facial expression dataset. This dataset contains soft-labels, three categories of data complexity subsets, and additional metadata such as age, gender, ethnicity, head pose, facial landmarks, valence, and arousal. AffectNet+ will be made publicly accessible to researchers.

Contact time of bouncing drops is one of the most essential parameters to quantify the water-repellency of surfaces. Generally, the contact time on superhydrophobic surfaces is known to be Weber number-independent. Here, we probe an additional characteristic time, \emph{transition time} inherent in water drop impacting on superhydrophobic surfaces, marking a switch from a predominantly lateral to an axial motion. Systematic experiments and numerical simulations show that the transition time is also Weber number-independent and accounts for half the contact time. Additionally we identify a Weber-independent partition of volume at the maximum spreading state between the rim and lamella and show that the latter contains 1/4 of the total volume of the drop.

With the arrival of the next generation of ultra-deep optical imaging surveys reaching $\mu_V$$\sim$30 mag/arcsec$^2$ (3$\sigma$; 10"$\times$10"), the removal of scattered light due to the point spread function (PSF) effect remains a critical step for the scientific exploitation of the low surface brightness information contained in these data. Because virtually all pixels in the ground-based images are affected by an unwanted screen of light with a brightness greater than $\mu_V$$\sim$29 mag/arcsec$^2$, the characterization of the extended PSF (R$>$5 arcmin) is mandatory. We describe the procedure used to construct the extended PSFs of the LIGHTS survey in the g- and r-band images taken with the Large Binocular Cameras (LBCs) of the Large Binocular Telescope (LBT). We produce PSFs with a radial extension of 7.5 arcmins. These are later extended to 30 arcmins following an empirically motivated power-law extrapolation of their behaviour in their outermost regions. As an example of the application of our methodology, we subtract the scattered light around the galaxy NGC3198. The result of this subtraction clearly shows the outermost parts of the galaxy's disc, which have been obscured by the influence of nearby bright stars. Following a reproducible science philosophy, we make all the PSF models and the scripts used to do the analysis of this paper publicly available., Comment: 13 pages, 14 figures, 1 tables

Over the past decades, improvements in data collection hardware coupled with novel artificial intelligence algorithms have made it possible for researchers to understand urban environments at an unprecedented scale. From local interactions between actors to city-wide infrastructural problems, this new data-driven approach enables a more informed and trustworthy decision-making process aiming at transforming cities into safer and more equitable places for living. This new moment unfolded new opportunities to understand various phenomena that directly impact how accessible cities are to heterogeneous populations. Specifically, sensing localized physical interactions among actors under different scenarios can drive substantial interventions in urban environments to make them safer for all. In this manuscript, we list opportunities and associated challenges to leverage street-level multimodal sensing data to empower domain experts in making more informed decisions and, ultimately, supporting a data-informed policymaking framework. The challenges presented here can motivate research in different areas, such as computer vision and human-computer interaction, to support cities in growing more sustainably., Comment: ASSETS 2024 UrbanAccess Workshop

The dominant paradigm for RLHF is online and on-policy RL: synchronously generating from the large language model (LLM) policy, labelling with a reward model, and learning using feedback on the LLM's own outputs. While performant, this paradigm is computationally inefficient. Inspired by classical deep RL literature, we propose separating generation and learning in RLHF. This enables asynchronous generation of new samples while simultaneously training on old samples, leading to faster training and more compute-optimal scaling. However, asynchronous training relies on an underexplored regime, online but off-policy RLHF: learning on samples from previous iterations of our model. To understand the challenges in this regime, we investigate a fundamental question: how much off-policyness can we tolerate for asynchronous training to speed up learning but maintain performance? Among several RLHF algorithms we tested, we find that online DPO is most robust to off-policy data, and robustness increases with the scale of the policy model. We study further compute optimizations for asynchronous RLHF but find that they come at a performance cost, giving rise to a trade-off. Finally, we verify the scalability of asynchronous RLHF by training LLaMA 3.1 8B on an instruction-following task 40% faster than a synchronous run while matching final performance., Comment: code at https://github.com/mnoukhov/async_rlhf

This booklet on Active Assisted Living (AAL) technologies has been created as part of the GoodBrother COST Action, which has run from 2020 to 2024. COST Actions are European research programs that promote collaboration across borders, uniting researchers, professionals, and institutions to address key societal challenges. GoodBrother focused on ethical and privacy concerns surrounding video and audio monitoring in care settings. The aim was to ensure that while AAL technologies help older adults and vulnerable individuals, their privacy and data protection rights remain a top priority. This booklet is designed to guide you through the role that AAL technologies play in improving the quality of life for older adults, caregivers, and people with disabilities. AAL technologies offer tools for those facing cognitive or physical challenges. They can enhance independence, assist with daily routines, and promote a safer living environment. However, the rise of these technologies also brings important questions about data protection and user autonomy. This resource is intended for a wide audience, including end users, caregivers, healthcare professionals, and policymakers. It provides practical guidance on integrating AAL technologies into care settings while safeguarding privacy and ensuring ethical use. The insights offered here aim to empower users and caregivers to make informed choices that enhance both the quality of care and respect for personal autonomy.

Recently, there have been numerous studies on feature learning with neural networks, specifically on learning single- and multi-index models where the target is a function of a low-dimensional projection of the input. Prior works have shown that in high dimensions, the majority of the compute and data resources are spent on recovering the low-dimensional projection; once this subspace is recovered, the remainder of the target can be learned independently of the ambient dimension. However, implications of feature learning in adversarial settings remain unexplored. In this work, we take the first steps towards understanding adversarially robust feature learning with neural networks. Specifically, we prove that the hidden directions of a multi-index model offer a Bayes optimal low-dimensional projection for robustness against $\ell_2$-bounded adversarial perturbations under the squared loss, assuming that the multi-index coordinates are statistically independent from the rest of the coordinates. Therefore, robust learning can be achieved by first performing standard feature learning, then robustly tuning a linear readout layer on top of the standard representations. In particular, we show that adversarially robust learning is just as easy as standard learning, in the sense that the additional number of samples needed to robustly learn multi-index models when compared to standard learning, does not depend on dimensionality., Comment: 39 pages, 1 figure

This paper presents Auriga Robotics' autonomous vehicle, developed at Shahid Beheshti University's Robotics and Intelligent Automation Lab, as part of the team's entry for the 2024 RoboCup IranOpen competition. The vehicle is a 1:10 scale car equipped with a custom-designed chassis, a stepper motor for precision, and a range of sensors for autonomous navigation. Key hardware includes ESP32 microcontrollers that manage motor control and sensor data acquisition. The software system integrates computer vision, including YOLOv8 for sign detection and PiNet for lane detection, combined with control algorithms such as the Stanley, PID, and Pure Pursuit controllers. The vehicle's design emphasizes real-time decision-making, environmental mapping, and efficient localization, ensuring its ability to navigate complex driving scenarios.

Decentralized Exchanges (DEXs) are pivotal applications in the Decentralized Finance (DeFi) landscape, aiming to facilitate trustless cryptocurrency trading by relying on smart contracts and blockchain networks. The developments in the DEXs sector began with the implementation of an Automated Market Maker (AMM) system using a simple math formula by Uniswap in 2018. Absorbing significant funding and the attention of web3 enthusiasts, DEXs have seen numerous advancements in their evolution. A notable recent advancement is the introduction of hooks in Uniswap v4, which allows users to take advantage of a wide range of plugin-like features with liquidity pools. This paper provides a comprehensive classification and comparative analyses of prominent DEX protocols, namely Uniswap, Curve, and Balancer, in addition to investigating other protocols' noteworthy aspects. The evaluation framework encompasses mechanisms, components, mathematical formulations, and the performance of liquidity pools. The goals are to elucidate the strengths and limitations of different AMM models, highlight emerging concepts in DEX development, outline current challenges, and differentiate optimal models for specific applications. The results and comparative insights can be a reference for web3 developers, blockchain researchers, traders, and regulatory parties.

We aim to discern scenarios of structural evolution of intermediate to high-mass star-forming galaxies (SFGs) since cosmic noon by comparing their stellar mass profiles with present-day stellar masses of $\log(M_{\ast,0}/M_{\odot})=10.3-11$. We addressed discrepancies in the size evolution rates of SFGs, which may be caused by variations in sample selection and methods for size measurements. To check these factors, we traced the evolution of individual galaxies by identifying their progenitors using stellar mass growth histories (SMGHs), integrating along the star-forming main sequence and from the IllustrisTNG simulations. Comparison between the structural parameters estimated from the mass- and light-based profiles shows that mass-weighted size evolves at a slower pace compared to light-based ones, highlighting the need to consider the mass-to-light ratio ($M/L$) gradients. Additionally, we observed mass-dependent growth in stellar mass profiles: massive galaxies ($\log(M_{\ast,0}/M_{\odot})\gtrsim10.8$) formed central regions at $z\gtrsim1.5$ and grew faster in outer regions, suggesting inside-out growth, while intermediate and less massive SFGs followed a relatively self-similar mass buildup since $z\sim2$. Moreover, slopes of observed size evolution conflict with the predictions of TNG50 for samples selected using the same SMGHs across our redshift range. To explore the origin of this deviation, we examined changes in angular momentum (AM) retention fraction using the half-mass size evolution and employing a simple disk formation model. Assuming similar dark matter halo parameters, our calculations indicate that the AM inferred from observations halved in the last 10 Gyr while it remained relatively constant in TNG50. This higher AM in simulations may be due to the accretion of high-AM gases into disks., Comment: Accepted for publication in ApJ; 25 pages, 11 figures

Classifiers trained with Empirical Risk Minimization (ERM) tend to rely on attributes that have high spurious correlation with the target. This can degrade the performance on underrepresented (or 'minority') groups that lack these attributes, posing significant challenges for both out-of-distribution generalization and fairness objectives. Many studies aim to enhance robustness to spurious correlation, but they sometimes depend on group annotations for training. Additionally, a common limitation in previous research is the reliance on group-annotated validation datasets for model selection. This constrains their applicability in situations where the nature of the spurious correlation is not known, or when group labels for certain spurious attributes are not available. To enhance model robustness with minimal group annotation assumptions, we propose Environment-based Validation and Loss-based Sampling (EVaLS). It uses the losses from an ERM-trained model to construct a balanced dataset of high-loss and low-loss samples, mitigating group imbalance in data. This significantly enhances robustness to group shifts when equipped with a simple post-training last layer retraining. By using environment inference methods to create diverse environments with correlation shifts, EVaLS can potentially eliminate the need for group annotation in validation data. In this context, the worst environment accuracy acts as a reliable surrogate throughout the retraining process for tuning hyperparameters and finding a model that performs well across diverse group shifts. EVaLS effectively achieves group robustness, showing that group annotation is not necessary even for validation. It is a fast, straightforward, and effective approach that reaches near-optimal worst group accuracy without needing group annotations, marking a new chapter in the robustness of trained models against spurious correlation.

Two-sided matching markets describe a large class of problems wherein participants from one side of the market must be matched to those from the other side according to their preferences. In many real-world applications (e.g. content matching or online labor markets), the knowledge about preferences may not be readily available and must be learned, i.e., one side of the market (aka agents) may not know their preferences over the other side (aka arms). Recent research on online settings has focused primarily on welfare optimization aspects (i.e. minimizing the overall regret) while paying little attention to the game-theoretic properties such as the stability of the final matching. In this paper, we exploit the structure of stable solutions to devise algorithms that improve the likelihood of finding stable solutions. We initiate the study of the sample complexity of finding a stable matching, and provide theoretical bounds on the number of samples needed to reach a stable matching with high probability. Finally, our empirical results demonstrate intriguing tradeoffs between stability and optimality of the proposed algorithms, further complementing our theoretical findings., Comment: Add comparison with ODA

In $f(T)$ gravity, the theory modifies the gravitational action by introducing a function of the torsion scalar $T$. This approach allows for a different treatment of gravity than general relativity, particularly in cosmological contexts. Dynamical system analysis is a powerful tool for exploring the stability and behavior of cosmological solutions within this framework. The dynamical system analysis involves examining the phase space of the cosmological equations derived from the $f(T)$ model. This analysis helps identify fixed points, stability, and the evolution of the universe's scale factor. Therefore, in the following, we first review the main equations of the $f(T)$ gravity model. Then we study the dynamic analysis of the gravity model and obtain stability points. Finally, we consider the bouncing scenario in this model., Comment: 5 pages, 2 figures

We study the depth of grade-school math (GSM) problem-solving capabilities of LLMs. To this end, we evaluate their performance on pairs of existing math word problems together so that the answer to the second problem depends on correctly answering the first problem. Our findings reveal a significant reasoning gap in most LLMs, that is performance difference between solving the compositional pairs and solving each question independently. This gap is more pronounced in smaller, more cost-efficient, and math-specialized models. Moreover, instruction-tuning recipes and code generation have varying effects across LLM sizes, while finetuning on GSM can lead to task overfitting. Our analysis indicates that large reasoning gaps are not because of test-set leakage, but due to distraction from additional context and poor second-hop reasoning. Overall, LLMs exhibit systematic differences in their reasoning abilities, despite what their performance on standard benchmarks indicates.

Detecting credit card fraud (CCF) holds significant importance due to its role in safeguarding consumers from unauthorized transactions that have the potential to result in financial detriment and negative impacts on their credit rating. It aids financial institutions in upholding the reliability of their payment mechanisms and circumventing the expensive procedure of compensating for deceitful transactions. The utilization of Artificial Intelligence methodologies demonstrated remarkable efficacy in the identification of credit card fraud instances. Within this study, we present a unique attention-based ensemble model. This model is enhanced by adding an attention layer for integration of first layer classifiers' predictions and a selection layer for choosing the best integrated value. The attention layer is implemented with two aggregation operators: dependent ordered weighted averaging (DOWA) and induced ordered weighted averaging (IOWA). The performance of the IOWA operator is very close to the learning algorithm in neural networks which is based on the gradient descent optimization method, and performing the DOWA operator is based on weakening the classifiers that make outlier predictions compared to other learners. Both operators have a sufficient level of complexity for the recognition of complex patterns. Accuracy and diversity are the two criteria we use for selecting the classifiers whose predictions are to be integrated by the two aggregation operators. Using a bootstrap forest, we identify the 13 most significant features of the dataset that contribute the most to CCF detection and use them to feed the proposed model. Exhibiting its efficacy, the ensemble model attains an accuracy of 99.95% with an area under the curve (AUC) of 1.

One of the most important challenges in the financial and cryptocurrency field is accurately predicting cryptocurrency price trends. Leveraging artificial intelligence (AI) is beneficial in addressing this challenge. Cryptocurrency markets, marked by substantial growth and volatility, attract investors and scholars keen on deciphering and forecasting cryptocurrency price movements. The vast and diverse array of data available for such predictions increases the complexity of the task. In our study, we introduce a novel approach termed hard and soft information fusion (HSIF) to enhance the accuracy of cryptocurrency price movement forecasts. The hard information component of our approach encompasses historical price records alongside technical indicators. Complementing this, the soft data component extracts from X (formerly Twitter), encompassing news headlines and tweets about the cryptocurrency. To use this data, we use the Bidirectional Encoder Representations from Transformers (BERT)-based sentiment analysis method, financial BERT (FinBERT), which performs best. Finally, our model feeds on the information set including processed hard and soft data. We employ the bidirectional long short-term memory (BiLSTM) model because processing information in both forward and backward directions can capture long-term dependencies in sequential information. Our empirical findings emphasize the superiority of the HSIF approach over models dependent on single-source data by testing on Bitcoin-related data. By fusing hard and soft information on Bitcoin dataset, our model has about 96.8\% accuracy in predicting price movement. Incorporating information enables our model to grasp the influence of social sentiment on price fluctuations, thereby supplementing the technical analysis-based predictions derived from hard information.

We study the contraction of strictly convex, axially symmetric hypersurfaces by a non-symmetric, non-homogeneous, fully nonlinear function of curvature. Starting from axially symmetric hypersurfaces with even profile curves, we show evolving hypersurfaces converge to a point in a finite time, and under proper rescaling, solutions will converge to a convex hypersurface.

With robots increasingly collaborating with humans in everyday tasks, it is important to take steps toward robotic systems capable of understanding the environment. This work focuses on scene understanding to detect pick and place tasks given initial and final images from the scene. To this end, a dataset is collected for object detection and pick and place task detection. A YOLOv5 network is subsequently trained to detect the objects in the initial and final scenes. Given the detected objects and their bounding boxes, two methods are proposed to detect the pick and place tasks which transform the initial scene into the final scene. A geometric method is proposed which tracks objects' movements in the two scenes and works based on the intersection of the bounding boxes which moved within scenes. Contrarily, the CNN-based method utilizes a Convolutional Neural Network to classify objects with intersected bounding boxes into 5 classes, showing the spatial relationship between the involved objects. The performed pick and place tasks are then derived from analyzing the experiments with both scenes. Results show that the CNN-based method, using a VGG16 backbone, outscores the geometric method by roughly 12 percentage points in certain scenarios, with an overall success rate of 84.3%., Comment: Conference Paper, ICEE 2024, 7 pages, 5 figures

One of the growing trends in machine learning is the use of data generation techniques, since the performance of machine learning models is dependent on the quantity of the training dataset. However, in many medical applications, collecting large datasets is challenging due to resource constraints, which leads to overfitting and poor generalization. This paper introduces a novel method, Artificial Data Point Generation in Clustered Latent Space (AGCL), designed to enhance classification performance on small medical datasets through synthetic data generation. The AGCL framework involves feature extraction, K-means clustering, cluster evaluation based on a class separation metric, and the generation of synthetic data points from clusters with distinct class representations. This method was applied to Parkinson's disease screening, utilizing facial expression data, and evaluated across multiple machine learning classifiers. Experimental results demonstrate that AGCL significantly improves classification accuracy compared to baseline, GN and kNNMTD. AGCL achieved the highest overall test accuracy of 83.33% and cross-validation accuracy of 90.90% in majority voting over different emotions, confirming its effectiveness in augmenting small datasets., Comment: 8 pages, 2 figures

Given two Deep Neural Network (DNN) classifiers with the same input and output domains, our goal is to quantify the robustness of the two networks in relation to each other. Towards this, we introduce the notion of Relative Safety Margins (RSMs). Intuitively, given two classes and a common input, RSM of one classifier with respect to another reflects the relative margins with which decisions are made. The proposed notion is relevant in the context of several applications domains, including to compare a trained network and its corresponding compact network (e.g., pruned, quantized, distilled network). Not only can RSMs establish whether decisions are preserved, but they can also quantify their qualities. We also propose a framework to establish safe bounds on RSM gains or losses given an input and a family of perturbations. We evaluate our approach using the MNIST, CIFAR10, and two real-world medical datasets, to show the relevance of our results.

We introduce a novel category of set functions called Extended Deep Submodular functions (EDSFs), which are neural network-representable. EDSFs serve as an extension of Deep Submodular Functions (DSFs), inheriting crucial properties from DSFs while addressing innate limitations. It is known that DSFs can represent a limiting subset of submodular functions. In contrast, through an analysis of polymatroid properties, we establish that EDSFs possess the capability to represent all monotone submodular functions, a notable enhancement compared to DSFs. Furthermore, our findings demonstrate that EDSFs can represent any monotone set function, indicating the family of EDSFs is equivalent to the family of all monotone set functions. Additionally, we prove that EDSFs maintain the concavity inherent in DSFs when the components of the input vector are non-negative real numbers-an essential feature in certain combinatorial optimization problems. Through extensive experiments, we illustrate that EDSFs exhibit significantly lower empirical generalization error than DSFs in the learning of coverage functions. This suggests that EDSFs present a promising advancement in the representation and learning of set functions with improved generalization capabilities.

Evaluating production-level retrieval systems at scale is a crucial yet challenging task due to the limited availability of a large pool of well-trained human annotators. Large Language Models (LLMs) have the potential to address this scaling issue and offer a viable alternative to humans for the bulk of annotation tasks. In this paper, we propose a framework for assessing the product search engines in a large-scale e-commerce setting, leveraging Multimodal LLMs for (i) generating tailored annotation guidelines for individual queries, and (ii) conducting the subsequent annotation task. Our method, validated through deployment on a large e-commerce platform, demonstrates comparable quality to human annotations, significantly reduces time and cost, facilitates rapid problem discovery, and provides an effective solution for production-level quality control at scale., Comment: 13 pages, 5 figures, 4 Tables

An image denoiser can be used for a wide range of restoration problems via the Plug-and-Play (PnP) architecture. In this paper, we propose a general framework to build an interpretable graph-based deep denoiser (GDD) by unrolling a solution to a maximum a posteriori (MAP) problem equipped with a graph Laplacian regularizer (GLR) as signal prior. Leveraging a recent theorem showing that any (pseudo-)linear denoiser $\boldsymbol \Psi$, under mild conditions, can be mapped to a solution of a MAP denoising problem regularized using GLR, we first initialize a graph Laplacian matrix $\mathbf L$ via truncated Taylor Series Expansion (TSE) of $\boldsymbol \Psi^{-1}$. Then, we compute the MAP linear system solution by unrolling iterations of the conjugate gradient (CG) algorithm into a sequence of neural layers as a feed-forward network -- one that is amenable to parameter tuning. The resulting GDD network is "graph-interpretable", low in parameter count, and easy to initialize thanks to $\mathbf L$ derived from a known well-performing denoiser $\boldsymbol \Psi$. Experimental results show that GDD achieves competitive image denoising performance compared to competitors, but employing far fewer parameters, and is more robust to covariate shift.

As autonomous vehicles become more prevalent, highly accurate and efficient systems are increasingly critical to improve safety, performance, and energy consumption. Efficient management of energy-reliability tradeoffs in these systems demands the ability to predict various conditions during vehicle operations. With the promising improvement of Large Language Models (LLMs) and the emergence of well-known models like ChatGPT, unique opportunities for autonomous vehicle-related predictions have been provided in recent years. This paper proposed MAPS using LLMs as map reader co-drivers to predict the vital parameters to set during the autonomous vehicle operation to balance the energy-reliability tradeoff. The MAPS method demonstrates a 20% improvement in navigation accuracy compared to the best baseline method. MAPS also shows 11% energy savings in computational units and up to 54% in both mechanical and computational units.

Automated emotion recognition has applications in various fields, such as human-machine interaction, healthcare, security, education, and emotion-aware recommendation/feedback systems. Developing methods to analyze human emotions accurately is essential to enable such diverse applications. Multiple studies have been conducted to explore the possibility of using physiological signals and machine-learning techniques to evaluate human emotions. Furthermore, internal factors such as personality have been considered and involved in emotion recognition. However, integrating personality that is user specific within traditional machine-learning methods that use user-agnostic large data sets has become a critical problem. This study proposes the APEX: attention on personality-based emotion recognition framework, in which multiple weak classifiers are trained on physiological signals of each participant's data, and the classification results are reweighed based on the personality correlations between corresponding subjects and test subjects. Experiments have been conducted on the ASCERTAIN dataset, and the results show that the proposed framework outperforms existing studies.

This paper presents a novel Electrodermal Activity (EDA) signal acquisition system, designed to address the challenges of stress monitoring in contemporary society, where stress affects one in four individuals. Our system focuses on enhancing the accuracy and efficiency of EDA measurements, a reliable indicator of stress. Traditional EDA monitoring solutions often grapple with trade-offs between sensor placement, cost, and power consumption, leading to compromised data accuracy. Our innovative design incorporates an adaptive gain mechanism, catering to the broad dynamic range and high-resolution needs of EDA data analysis. The performance of our system was extensively tested through simulations and a custom Printed Circuit Board (PCB), achieving an error rate below 1\% and maintaining power consumption at a mere 700$\mu$A under a 3.7V power supply. This research contributes significantly to the field of wearable health technology, offering a robust and efficient solution for long-term stress monitoring.

Brain tumors are the most common solid tumors in children and young adults, but the scarcity of large histopathology datasets has limited the application of computational pathology in this group. This study implements two weakly supervised multiple-instance learning (MIL) approaches on patch-features obtained from state-of-the-art histology-specific foundation models to classify pediatric brain tumors in hematoxylin and eosin whole slide images (WSIs) from a multi-center Swedish cohort. WSIs from 540 subjects (age 8.5$\pm$4.9 years) diagnosed with brain tumor were gathered from the six Swedish university hospitals. Instance (patch)-level features were obtained from WSIs using three pre-trained feature extractors: ResNet50, UNI and CONCH. Instances were aggregated using attention-based MIL (ABMIL) or clustering-constrained attention MIL (CLAM) for patient-level classification. Models were evaluated on three classification tasks based on the hierarchical classification of pediatric brain tumors: tumor category, family and type. Model generalization was assessed by training on data from two of the centers and testing on data from four other centers. Model interpretability was evaluated through attention-mapping. The highest classification performance was achieved using UNI features and AMBIL aggregation, with Matthew's correlation coefficient of 0.86$\pm$0.04, 0.63$\pm$0.04, and 0.53$\pm$0.05, for tumor category, family and type classification, respectively. When evaluating generalization, models utilizing UNI and CONCH features outperformed those using ResNet50. However, the drop in performance from the in-site to out-of-site testing was similar across feature extractors. These results show the potential of state-of-the-art computational pathology methods in diagnosing pediatric brain tumors at different hierarchical levels with fair generalizability on a multi-center national dataset.

Radiological images, such as magnetic resonance imaging (MRI) and computed tomography (CT) images, typically consist of a body part and a dark background. For many analyses, it is necessary to separate the body part from the background. In this article, we present a Python code designed to separate body and background regions in 2D and 3D radiological images. We tested the algorithm on various MRI and CT images of different body parts, including the brain, neck, and abdominal regions. Additionally, we introduced a method for intensity normalization and outlier restriction, adjusted for data conversion into 8-bit unsigned integer (UINT8) format, and examined its effects on body-background separation. Our Python code is available for use with proper citation., Comment: 14 pages, 8 figures. typos corrected

Training on high-quality synthetic data from strong language models (LMs) is a common strategy to improve the reasoning performance of LMs. In this work, we revisit whether this strategy is compute-optimal under a fixed inference budget (e.g., FLOPs). To do so, we investigate the trade-offs between generating synthetic data using a stronger but more expensive (SE) model versus a weaker but cheaper (WC) model. We evaluate the generated data across three key metrics: coverage, diversity, and false positive rate, and show that the data from WC models may have higher coverage and diversity, but also exhibit higher false positive rates. We then finetune LMs on data from SE and WC models in different settings: knowledge distillation, self-improvement, and a novel weak-to-strong improvement setup where a weaker LM teaches reasoning to a stronger LM. Our findings reveal that models finetuned on WC-generated data consistently outperform those trained on SE-generated data across multiple benchmarks and multiple choices of WC and SE models. These results challenge the prevailing practice of relying on SE models for synthetic data generation, suggesting that WC may be the compute-optimal approach for training advanced LM reasoners.

Verifiers or reward models are often used to enhance the reasoning performance of large language models (LLMs). A common approach is the Best-of-N method, where N candidate solutions generated by the LLM are ranked by a verifier, and the best one is selected. While LLM-based verifiers are typically trained as discriminative classifiers to score solutions, they do not utilize the text generation capabilities of pretrained LLMs. To overcome this limitation, we instead propose training verifiers using the ubiquitous next-token prediction objective, jointly on verification and solution generation. Compared to standard verifiers, such generative verifiers (GenRM) can benefit from several advantages of LLMs: they integrate seamlessly with instruction tuning, enable chain-of-thought reasoning, and can utilize additional test-time compute via majority voting for better verification. We demonstrate that GenRM outperforms discriminative, DPO verifiers, and LLM-as-a-Judge, resulting in a 16-40% improvement in the number of problems solved with Best-of-N on algorithmic and math reasoning tasks. Furthermore, we find that training GenRM with synthetic verification rationales is sufficient to pick out subtle errors on math problems. Finally, we demonstrate that generative verifiers scale favorably with model size and inference-time compute.

We develop a model describing long-range atom-atom interactions in a two-dimensional periodic or aperiodic lattice of optical centers inside a solid-state host material. We consider realistic environmental and technical conditions such as frequency and position broadening. Even when considering a significant frequency broadening in the ensemble (approximately 300 GHz), we observe up to a three-fold increase in directional scattering from the resonant lattice in a system. The model can be used to scalably design quantum optical elements, e.g. a quantum lens, harnessing atomistic engineering (e.g. via ion implantation) of collective interactions in materials to enhance quantum properties., Comment: 15 pages, 10 figures