Search

One of the main goals of gravitational-wave astrophysics is to study gravity in the strong-field regime and constrain deviations from general relativity. Any such deviation affects not only binary dynamics and gravitational-wave emission but also the structure and tidal properties of compact objects. In the case of neutron stars, masses, radii, and tidal deformabilities can all differ significantly between different theories of gravity. Currently, the measurement uncertainties in neutron-star radii and tidal deformabilities are quite large. However, much less is known about how the large uncertainty in the nuclear equation of state might affect tests of general relativity using binary neutron-star mergers. Conversely, using the wrong theory of gravity might lead to incorrect constraints on the nuclear equation of state. Here we study this problem within scalar-tensor theory. We apply the recently derived $\ell = 2$ tidal love numbers in this theory to parameter estimation of GW170817. Correspondingly, we test if physics beyond general relativity could bias measurements of the nuclear equation of state and neutron-star radii. We find that parameter inference for both the general relativistic and scalar-tensor case return consistent component masses and tidal deformabilites. The radius and the equation of state posteriors, however, differ between the two theories, but neither is excluded by current observational limits. This indicates that measurements of the nuclear equation of state may be biased and that deviations from general relativity could go undetected when analyzing current binary neutron star mergers., Comment: 9 pages. 2 figures. To be submitted to APJL. Comments welcome

Large language models (LLMs), trained on diverse data effectively acquire a breadth of information across various domains. However, their computational complexity, cost, and lack of transparency hinder their direct application for specialised tasks. In fields such as clinical research, acquiring expert annotations or prior knowledge about predictive models is often costly and time-consuming. This study proposes using LLMs to elicit expert prior distributions for predictive models. This approach also provides an alternative to in-context learning, where language models are tasked with making predictions directly. We compare LLM-elicited and uninformative priors, evaluate whether LLMs truthfully generate parameter distributions, and propose a model selection strategy for in-context learning and prior elicitation. Our findings show that LLM-elicited prior parameter distributions significantly reduce predictive error compared to uninformative priors in low-data settings. Applied to clinical problems, this translates to fewer required biological samples, lowering cost and resources. Prior elicitation also consistently outperforms and proves more reliable than in-context learning at a lower cost, making it a preferred alternative in our setting. We demonstrate the utility of this method across various use cases, including clinical applications. For infection prediction, using LLM-elicited priors reduced the number of required labels to achieve the same accuracy as an uninformative prior by 55%, at 200 days earlier in the study.

Recent advancements in artificial intelligence have sparked interest in scientific assistants that could support researchers across the full spectrum of scientific workflows, from literature review to experimental design and data analysis. A key capability for such systems is the ability to process and reason about scientific information in both visual and textual forms - from interpreting spectroscopic data to understanding laboratory setups. Here, we introduce MaCBench, a comprehensive benchmark for evaluating how vision-language models handle real-world chemistry and materials science tasks across three core aspects: data extraction, experimental understanding, and results interpretation. Through a systematic evaluation of leading models, we find that while these systems show promising capabilities in basic perception tasks - achieving near-perfect performance in equipment identification and standardized data extraction - they exhibit fundamental limitations in spatial reasoning, cross-modal information synthesis, and multi-step logical inference. Our insights have important implications beyond chemistry and materials science, suggesting that developing reliable multimodal AI scientific assistants may require advances in curating suitable training data and approaches to training those models.

Checklists have been widely recognized as effective tools for completing complex tasks in a systematic manner. Although originally intended for use in procedural tasks, their interpretability and ease of use have led to their adoption for predictive tasks as well, including in clinical settings. However, designing checklists can be challenging, often requiring expert knowledge and manual rule design based on available data. Recent work has attempted to address this issue by using machine learning to automatically generate predictive checklists from data, although these approaches have been limited to Boolean data. We propose a novel method for learning predictive checklists from diverse data modalities, such as images and time series. Our approach relies on probabilistic logic programming, a learning paradigm that enables matching the discrete nature of checklist with continuous-valued data. We propose a regularization technique to tradeoff between the information captured in discrete concepts of continuous data and permit a tunable level of interpretability for the learned checklist concepts. We demonstrate that our method outperforms various explainable machine learning techniques on prediction tasks involving image sequences, time series, and clinical notes., Comment: 36 pages

In 1+1 dimensions, the equations of motion of the Horowitz-Polchinski (HP) effective string have a re-writing in terms of a first order system. This is attributed to FZZ duality. In this note, we observe that a similar re-writing exists in higher dimensions, so that the degree of the dilaton-winding subsystem reduces to first order. The 1+1 first order equations emerge as a natural limit of the higher dimensional HP system in the cap region of the cigar. As a result, there is a critical value of the winding amplitude that matches with the 1+1 coset SCFT prediction. At this critical point, the cigar has a puncture at the Euclidean horizon and the $higher$ $dimensional$ black hole entropy is correctly reproduced by the winding condensate., Comment: 25 pp

This paper presents the results of a joint radio and gamma-ray timing campaign on the nine millisecond pulsars (MSPs) discovered as part of the L-band targeted survey of Fermi-LAT sources performed in the context of the Transients and Pulsars with MeerKAT (TRAPUM) Large Survey Project. Out of these pulsars, eight are members of binary systems; of these eight, two exhibit extended eclipses of the radio emission. Using an initial radio timing solution, pulsations were found in the gamma rays for six of the targets. For these sources, a joint timing analysis of radio times of arrival and gamma-ray photons was performed, using a newly developed code that optimises the parameters through a Markov chain Monte Carlo (MCMC) technique. This approach has allowed us to precisely measure both the short- and long-term timing parameters. This study includes a proper motion measurement for four pulsars, which a gamma ray-only analysis would not have been sensitive to, despite the 15-year span of Fermi data., Comment: 20 pages, 5 figures

A dip in coincidence peaks for an electron beam is an experimental signature to detect Coulomb repulsion and Pauli pressure. This paper discusses another effect that can produce a similar signature but that does not originate from the properties of the physical system under scrutiny. Instead, the detectors and electronics used to measure those coincidences suffer significantly even from weak crosstalk. A simple model that explains our experimental observations is given. Furthermore we provide an experimental approach to correct for this type of crosstalk., Comment: 6 pages, 4 figures

The proliferation of big data and analytic workloads has driven the need for cloud compute and cluster-based job processing. With Apache Spark, users can process terabytes of data at ease with hundreds of parallel executors. At Microsoft, we aim at providing a fast and succinct interface for users to run Spark applications, such as through creating simple notebook "sessions" by abstracting the underlying complexity of the cloud. Providing low latency access to Spark clusters and sessions is a challenging problem due to the large overheads of cluster creation and session startup. In this paper, we introduce Intelligent Pooling, a system for proactively provisioning compute resources to combat the aforementioned overheads. To reduce the COGS (cost-of-goods-sold), our system (1) predicts usage patterns using an innovative hybrid Machine Learning (ML) model with low latency and high accuracy; and (2) optimizes the pool size dynamically to meet customer demand while reducing extraneous COGS. The proposed system auto-tunes its hyper-parameters to balance between performance and operational cost with minimal to no engineering input. Evaluated using large-scale production data, Intelligent Pooling achieves up to 43% reduction in cluster idle time compared to static pooling when targeting 99% pool hit rate. Currently deployed in production, Intelligent Pooling is on track to save tens of million dollars in COGS per year as compared to traditional pre-provisioned pools.

Cloud operators have expanded their service offerings, known as Stock Keeping Units (SKUs), to accommodate diverse demands, resulting in increased complexity for customers to select appropriate configurations. In a studied system, only 43% of the resource capacity was correctly chosen. Automated solutions addressing this issue often require enriched data, such as workload traces, which are unavailable for new services. However, telemetry from existing users and customer satisfaction feedback provide valuable insights for understanding customer needs and improving provisioning recommendations. This paper introduces Lorentz, an intelligent SKU recommender for provisioning compute resources without relying on workload traces. Lorentz uses customer profile data to forecast resource capacities for new users by profiling existing ones. It also incorporates a continuous feedback loop to refine recommendations based on customer performance versus cost preferences inferred from satisfaction signals. Validated with production data from Azure PostgreSQL DB, Lorentz achieves over 60% slack reduction without increasing throttling compared to user selections and existing defaults. Evaluations with synthetic data demonstrate Lorentz's ability to iteratively learn user preferences with high accuracy.

Motivation: A major challenge in metabolomics is annotation: assigning molecular structures to mass spectral fragmentation patterns. Despite recent advances in molecule-to-spectra and in spectra-to-molecular fingerprint prediction (FP), annotation rates remain low. Results: We introduce in this paper a novel paradigm (JESTR) for annotation. Unlike prior approaches that explicitly construct molecular fingerprints or spectra, JESTR leverages the insight that molecules and their corresponding spectra are views of the same data and effectively embeds their representations in a joint space. Candidate structures are ranked based on cosine similarity between the embeddings of query spectrum and each candidate. We evaluate JESTR against mol-to-spec and spec-to-FP annotation tools on three datasets. On average, for rank@[1-5], JESTR outperforms other tools by 23.6%-71.6%. We further demonstrate the strong value of regularization with candidate molecules during training, boosting rank@1 performance by 11.4% and enhancing the model's ability to discern between target and candidate molecules. Through JESTR, we offer a novel promising avenue towards accurate annotation, therefore unlocking valuable insights into the metabolome., Comment: 10 pages, 10 figures, 4 tables

As AI systems become integral to critical operations across industries and services, ensuring their reliability and safety is essential. We offer a framework that integrates established reliability and resilience engineering principles into AI systems. By applying traditional metrics such as failure rate and Mean Time Between Failures (MTBF) along with resilience engineering and human reliability analysis, we propose an integrate framework to manage AI system performance, and prevent or efficiently recover from failures. Our work adapts classical engineering methods to AI systems and outlines a research agenda for future technical studies. We apply our framework to a real-world AI system, using system status data from platforms such as openAI, to demonstrate its practical applicability. This framework aligns with emerging global standards and regulatory frameworks, providing a methodology to enhance the trustworthiness of AI systems. Our aim is to guide policy, regulation, and the development of reliable, safe, and adaptable AI technologies capable of consistent performance in real-world environments.

We investigate the variation in the total number of points in a random $p\times p$ square in $\mathbb{Z}^2$ where the $p$-adic valuation of a given polynomial in two variables is precisely $1$. We establish that this quantity follows a Poisson distribution as $p\rightarrow\infty$ under a certain conjecture. We also relate this conjecture to certain uniform distribution properties of a vector valued sequence., Comment: fixed typos

Recent advances in boundary critical phenomena have led to the discovery of a new surface universality class in the three-dimensional $O(N)$ model. The newly found "extraordinary-log" phase can be realized on a two-dimensional surface for $N< N_c$, with $N_c>3$, and on a plane defect embedded into a three-dimensional system, for any $N$. One of the key features of the extraordinary-log phase is the presence of logarithmic violations of standard finite-size scaling. In this work we study finite-size scaling in the extraordinary-log universality class by means of Monte Carlo simulations of an improved lattice model. We simulate the model with open boundary conditions, realizing the extraordinary-log phase on the surface for $N=2,3$, as well as with fully periodic boundary conditions and in the presence of a plane defect for $N=2,3,4$. In line with theory predictions, renormalization-group invariant observables studied here exhibit a logarithmic dependence on the size of the system. We numerically access not only the leading term in the $\beta$-function governing these logarithmic violations, but also the subleading term, which controls the evolution of the boundary phase diagram as a function of $N$., Comment: 18 pages, 11 figures

Transformers are set to become ubiquitous with applications ranging from chatbots and educational assistants to visual recognition and remote sensing. However, their increasing computational and memory demands is resulting in growing energy consumption. Building models with fast and energy-efficient inference is imperative to enable a variety of transformer-based applications. Look Up Table (LUT) based Weightless Neural Networks are faster than the conventional neural networks as their inference only involves a few lookup operations. Recently, an approach for learning LUT networks directly via an Extended Finite Difference method was proposed. We build on this idea, extending it for performing the functions of the Multi Layer Perceptron (MLP) layers in transformer models and integrating them with transformers to propose Quasi Weightless Transformers (QuWeiT). This allows for a computational and energy-efficient inference solution for transformer-based models. On I-ViT-T, we achieve a comparable accuracy of 95.64% on CIFAR-10 dataset while replacing approximately 55% of all the multiplications in the entire model and achieving a 2.2x energy efficiency. We also observe similar savings on experiments with the nanoGPT framework.

Diabetic Retinopathy DR is a severe complication of diabetes. Damaged or abnormal blood vessels can cause loss of vision. The need for massive screening of a large population of diabetic patients has generated an interest in a computer-aided fully automatic diagnosis of DR. In the realm of Deep learning frameworks, particularly convolutional neural networks CNNs, have shown great interest and promise in detecting DR by analyzing retinal images. However, several challenges have been faced in the application of deep learning in this domain. High-quality, annotated datasets are scarce, and the variations in image quality and class imbalances pose significant hurdles in developing a dependable model. In this paper, we demonstrate the proficiency of two Convolutional Neural Networks CNNs based models, UNET and Stacked UNET utilizing the APTOS Asia Pacific Tele-Ophthalmology Society Dataset. This system achieves an accuracy of 92.81% for the UNET and 93.32% for the stacked UNET architecture. The architecture classifies the images into five categories ranging from 0 to 4, where 0 is no DR and 4 is proliferative DR.

A crucial question in active patient care is determining if a treatment is having the desired effect, especially when changes are subtle over short periods. We propose using inter-patient data to train models that can learn to detect these fine-grained changes within a single patient. Specifically, can a model trained on multi-patient scans predict subtle changes in an individual patient's scans? Recent years have seen increasing use of deep learning (DL) in predicting diseases using biomedical imaging, such as predicting COVID-19 severity using lung ultrasound (LUS) data. While extensive literature exists on successful applications of DL systems when well-annotated large-scale datasets are available, it is quite difficult to collect a large corpus of personalized datasets for an individual. In this work, we investigate the ability of recent computer vision models to learn fine-grained differences while being trained on data showing larger differences. We evaluate on an in-house LUS dataset and a public ADNI brain MRI dataset. We find that models pre-trained on clips from multiple patients can better predict fine-grained differences in scans from a single patient by employing contrastive learning., Comment: Under review at ISBI 2025 conference

Electrostatics in the solution phase is governed by free electrical charges such as ions, as well as by bound charges that arise when a polarizable medium responds to an applied field. In a local medium, described by a constant dielectric permittivity, the sign of the far-field electrostatic potential distribution around an object is governed by its electrical charge. We demonstrate significant departures from this expectation in a nonlocal medium characterized by a wave vector-dependent dielectric function. Here, surface polarization due to the solvent, or indeed non-solvent dipoles, may wield significant influence at large distances. The polarization correlation length may not only significantly augment the effective screening length, but we show that the electrical contribution from polarization can compete with and even invert the sign of the electrical potential and the field arising from charge alone. These results hold ramifications for a range of apparently anomalous electrically governed observations such as underscreening, electrophoretic mobilities of charge-neutral objects, and long-ranged attraction between like-charged entities in water and other solvents.

This work explores conditions under which multi-finger grasping algorithms can attain robust sim-to-real transfer. While numerous large datasets facilitate learning generative models for multi-finger grasping at scale, reliable real-world dexterous grasping remains challenging, with most methods degrading when deployed on hardware. An alternate strategy is to use discriminative grasp evaluation models for grasp selection and refinement, conditioned on real-world sensor measurements. This paradigm has produced state-of-the-art results for vision-based parallel-jaw grasping, but remains unproven in the multi-finger setting. In this work, we find that existing datasets and methods have been insufficient for training discriminitive models for multi-finger grasping. To train grasp evaluators at scale, datasets must provide on the order of millions of grasps, including both positive and negative examples, with corresponding visual data resembling measurements at inference time. To that end, we release a new, open-source dataset of 3.5M grasps on 4.3K objects annotated with RGB images, point clouds, and trained NeRFs. Leveraging this dataset, we train vision-based grasp evaluators that outperform both analytic and generative modeling-based baselines on extensive simulated and real-world trials across a diverse range of objects. We show via numerous ablations that the key factor for performance is indeed the evaluator, and that its quality degrades as the dataset shrinks, demonstrating the importance of our new dataset. Project website at: https://sites.google.com/view/get-a-grip-dataset.

The spin-valley or Kramers qubit promises significantly enhanced spin-valley lifetimes due to strong coupling of the electrons' spin to their momentum (valley) degrees of freedom. In transition metal dichalcogenides (TMDCs) such spin-valley locking is expected to be particularly strong owing to the significant intrinsic spin-orbit coupling strength. Very recently, a small number of experiments on TMDC quantum dots have put forth evidence for spin-valley locking for the first time at the few-electron limit. Employing quantum transport theory, here we numerically simulate their ground- and excited-state transport spectroscopy signatures in a unified theoretical framework. In doing so, we reveal the operating conditions under which spin-valley locking occurs in TMDC quantum dots, thereby weaving the connection between intrinsic material properties and the experimental data under diverse conditions. Our simulations thus provide a predictive modeling tool for TMDC quantum dots at the few-electron limit allowing us to deduce from experiments the degree of spin-valley locking based on the SOC strength, inter-valley mixing, and the spin and valley $g$-factors. Our theoretical analysis provides an important milestone towards the next challenge of experimentally confirming valley-relaxation times using single-shot projective measurements, Comment: 13 pages, 6 figures with Appendix included, comments welcome

Estimated brain age from magnetic resonance image (MRI) and its deviation from chronological age can provide early insights into potential neurodegenerative diseases, supporting early detection and implementation of prevention strategies. Diffusion MRI (dMRI), a widely used modality for brain age estimation, presents an opportunity to build an earlier biomarker for neurodegenerative disease prediction because it captures subtle microstructural changes that precede more perceptible macrostructural changes. However, the coexistence of macro- and micro-structural information in dMRI raises the question of whether current dMRI-based brain age estimation models are leveraging the intended microstructural information or if they inadvertently rely on the macrostructural information. To develop a microstructure-specific brain age, we propose a method for brain age identification from dMRI that minimizes the model's use of macrostructural information by non-rigidly registering all images to a standard template. Imaging data from 13,398 participants across 12 datasets were used for the training and evaluation. We compare our brain age models, trained with and without macrostructural information minimized, with an architecturally similar T1-weighted (T1w) MRI-based brain age model and two state-of-the-art T1w MRI-based brain age models that primarily use macrostructural information. We observe difference between our dMRI-based brain age and T1w MRI-based brain age across stages of neurodegeneration, with dMRI-based brain age being older than T1w MRI-based brain age in participants transitioning from cognitively normal (CN) to mild cognitive impairment (MCI), but younger in participants already diagnosed with Alzheimer's disease (AD). Approximately 4 years before MCI diagnosis, dMRI-based brain age yields better performance than T1w MRI-based brain ages in predicting transition from CN to MCI.

The rapid expansion of foundation models (FMs), such as large language models (LLMs), has given rise to FMware--software systems that integrate FMs as core components. While building demonstration-level FMware is relatively straightforward, transitioning to production-ready systems presents numerous challenges, including reliability, high implementation costs, scalability, and compliance with privacy regulations. This paper provides a thematic analysis of the key obstacles in productionizing FMware, synthesized from industry experience and diverse data sources, including hands-on involvement in the Open Platform for Enterprise AI (OPEA) and FMware lifecycle engineering. We identify critical issues in FM selection, data and model alignment, prompt engineering, agent orchestration, system testing, and deployment, alongside cross-cutting concerns such as memory management, observability, and feedback integration. We discuss needed technologies and strategies to address these challenges and offer guidance on how to enable the transition from demonstration systems to scalable, production-ready FMware solutions. Our findings underscore the importance of continued research and multi-industry collaboration to advance the development of production-ready FMware.

The interstellar medium of the Milky Way contains turbulent plasma with structures driven by energetic processes that fuel star formation and shape the evolution of our Galaxy. Radio waves from pulsars are scattered off the small (au-scale and below) structures, resulting in frequency-dependent interference patterns that are modulated in time because of the relative motions of the pulsar, Earth, and plasma. Power spectral analyses of these patterns show parabolic arcs with curvatures that encode the locations and kinematics of individual structures. Here we report the discovery of at least 25 distinct plasma structures in the direction of the brilliant millisecond pulsar, PSR J0437$-$4715, in observations obtained with the MeerKAT radio telescope. Four arcs reveal structures within 5000 au of the pulsar, from a series of shocks induced as the pulsar and its wind interact with the ambient insterstellar medium. The measured radial distance and velocity of the main shock allows us to solve the shock geometry and space velocity of the pulsar in three dimensions, while the velocity of another structure unexpectedly indicates a back flow from the direction of the shock or pulsar-wind tail. The remaining 21 arcs represent a surprising abundance of structures sustained by turbulence within the Local Bubble -- a region of the interstellar medium thought to be depleted of gas by a series of supernova explosions about 14 Myr ago., Comment: 46 pages, 10 figures, 1 table, submitted to Nature Astronomy

Friends-and-strangers graphs, coined by Defant and Kravitz, are denoted by $\mathsf{FS}(X,Y)$ where $X$ and $Y$ are both graphs on $n$ vertices. The graph $X$ represents positions and edges mark adjacent positions while the graph $Y$ represents people and edges mark friendships. The vertex set of $\mathsf{FS}(X,Y)$ consists of all one-to-one placements of people on positions, and there is an edge between any two placements if it is possible to swap two people who are friends and on adjacent positions to get from one placement to the other. Previous papers have studied when $\mathsf{FS}(X,Y)$ is connected. In this paper, we consider when $\mathsf{FS}(X,Y)$ is $k$-connected where a graph is $k$-connected if it remains connected after removing any $k-1$ or less vertices. We first consider $\mathsf{FS}(X,Y)$ when $Y$ is a complete graph or star graph. We find tight bounds on their connectivity, proving their connectivity equals their minimum degree. We further consider the size of the connected components of $\mathsf{FS}(X,\mathsf{Star}_n)$ where $X$ is connected. We show that asymptotically similar conditions as the conditions mentioned by Bangachev are sufficient for $\mathsf{FS}(X,Y)$ to be $k$-connected. Finally, we consider when $X$ and $Y$ are independent Erd\H{o}s--R\'enyi random graphs on $n$ vertices and edge probability $p_1$ and $p_2,$ respectively. We show that for $p_0 = n^{-1/2+o(1)},$ if $p_1p_2\geq p_0^2$ and $p_1,$ $p_2 \geq w(n) p_0$ where $w(n) \rightarrow 0$ as $n \rightarrow \infty,$ then $\mathsf{FS}(X,Y)$ is $k$-connected with high probability. This is asymptotically tight as we show that below an asymptotically similar threshold $p_0'=n^{-1/2+o(1)}$, the graph $\mathsf{FS}(X,Y)$ is disconnected with high probability if $p_1p_2 \leq (p_0')^2$., Comment: 35 pages, 9 figures

Effective communication within universities is crucial for addressing the diverse information needs of students, alumni, and external stakeholders. However, existing chatbot systems often fail to deliver accurate, context-specific responses, resulting in poor user experiences. In this paper, we present KatzBot, an innovative chatbot powered by KatzGPT, a custom Large Language Model (LLM) fine-tuned on domain-specific academic data. KatzGPT is trained on two university-specific datasets: 6,280 sentence-completion pairs and 7,330 question-answer pairs. KatzBot outperforms established existing open source LLMs, achieving higher accuracy and domain relevance. KatzBot offers a user-friendly interface, significantly enhancing user satisfaction in real-world applications. The source code is publicly available at \url{https://github.com/AiAI-99/katzbot}.

Artificial neural networks (ANNs) show limited performance with scarce or imbalanced training data and face challenges with continuous learning, such as forgetting previously learned data after new tasks training. In contrast, the human brain can learn continuously and from just a few examples. This research explores the impact of 'sleep', an unsupervised phase incorporating stochastic activation with local Hebbian learning rules, on ANNs trained incrementally with limited and imbalanced datasets, specifically MNIST and Fashion MNIST. We discovered that introducing a sleep phase significantly enhanced accuracy in models trained with limited data. When a few tasks were trained sequentially, sleep replay not only rescued previously learned information that had been catastrophically forgetting following new task training but often enhanced performance in prior tasks, especially those trained with limited data. This study highlights the multifaceted role of sleep replay in augmenting learning efficiency and facilitating continual learning in ANNs.

We design a quantum thermal device that can simultaneously and dynamically cool multiple target qubits. Using a setup with three bosonic heat baths, we propose an engineering of interaction Hamiltonian using operators on different subspaces of the full Hilbert space of the system labelled by different magnetizations. We demonstrate, using the local as well as global quantum master equations, that a set of target qubits can be cooled simultaneously using these interaction Hamiltonians, while equal cooling of all target qubits is possible only when the local quantum master equation is used. However, the amount of cooling obtained from different magnetization subspaces, as quantified by a distance-based measure of qubit-local steady-state temperatures, may vary. We also investigate cooling of a set of target qubits when the interaction Hamiltonian has different magnetization components, and when the design of the quantum thermal device involves two heat baths instead of three. Further, we demonstrate, using local quantum master equation, that during providing cooling to the target qubits, the designed device operates only as a quantum absorption refrigerator. In contrast, use of the global quantum master equation indicates cooling of the target qubits even when the device works outside the operation regime of a quantum absorption refrigerator. We also extend the design to a star network of qubits interacting via Heisenberg interaction among each other, kept in contact with either three, or two heat baths, and discuss cooling of a set of target qubits using this device., Comment: 19 pages, 9 figures

Reinforcement Learning from Human Feedback (RLHF) is widely used to align Language Models (LMs) with human preferences. However, existing approaches often neglect individual user preferences, leading to suboptimal personalization. We present the Preference Pretrained Transformer (PPT), a novel approach for adaptive personalization using online user feedback. PPT leverages the in-context learning capabilities of transformers to dynamically adapt to individual preferences. Our approach consists of two phases: (1) an offline phase where we train a single policy model using a history-dependent loss function, and (2) an online phase where the model adapts to user preferences through in-context learning. We demonstrate PPT's effectiveness in a contextual bandit setting, showing that it achieves personalized adaptation superior to existing methods while significantly reducing the computational costs. Our results suggest the potential of in-context learning for scalable and efficient personalization in large language models.

Large Language Models excel in tasks like natural language understanding and text generation. Prompt engineering plays a critical role in leveraging LLM effectively. However, LLMs black-box nature hinders its interpretability and effective prompting engineering. A wide range of model explanation approaches have been developed for deep learning models, However, these local explanations are designed for single-output tasks like classification and regression,and cannot be directly applied to LLMs, which generate sequences of tokens. Recent efforts in LLM explanation focus on natural language explanations, but they are prone to hallucinations and inaccuracies. To address this, we introduce PromptExp , a framework for multi-granularity prompt explanations by aggregating token-level insights. PromptExp introduces two token-level explanation approaches: 1. an aggregation-based approach combining local explanation techniques, and 2. a perturbation-based approach with novel techniques to evaluate token masking impact. PromptExp supports both white-box and black-box explanations and extends explanations to higher granularity levels, enabling flexible analysis. We evaluate PromptExp in case studies such as sentiment analysis, showing the perturbation-based approach performs best using semantic similarity to assess perturbation impact. Furthermore, we conducted a user study to confirm PromptExp's accuracy and practical value, and demonstrate its potential to enhance LLM interpretability., Comment: 11 pages

This paper presents a method for generating large-scale datasets to improve class-agnostic video segmentation across robots with different form factors. Specifically, we consider the question of whether video segmentation models trained on generic segmentation data could be more effective for particular robot platforms if robot embodiment is factored into the data generation process. To answer this question, a pipeline is formulated for using 3D reconstructions (e.g. from HM3DSem) to generate segmented videos that are configurable based on a robot's embodiment (e.g. sensor type, sensor placement, and illumination source). A resulting massive RGB-D video panoptic segmentation dataset (MVPd) is introduced for extensive benchmarking with foundation and video segmentation models, as well as to support embodiment-focused research in video segmentation. Our experimental findings demonstrate that using MVPd for finetuning can lead to performance improvements when transferring foundation models to certain robot embodiments, such as specific camera placements. These experiments also show that using 3D modalities (depth images and camera pose) can lead to improvements in video segmentation accuracy and consistency. The project webpage is available at https://topipari.com/projects/MVPd, Comment: Accepted in IEEE Robotics and Automation Letters October 2024

Patterns arise spontaneously in a range of systems spanning the sciences, and their study typically focuses on mechanisms to understand their evolution in space-time. Increasingly, there has been a transition towards controlling these patterns in various functional settings, with implications for engineering. Here, we combine our knowledge of a general class of dynamical laws for pattern formation in non-equilibrium systems, and the power of stochastic optimal control approaches to present a framework that allows us to control patterns at multiple scales, which we dub the "Hamiltonian bridge". We use a mapping between stochastic many-body Lagrangian physics and deterministic Eulerian pattern forming PDEs to leverage our recent approach utilizing the Feynman-Kac-based adjoint path integral formulation for the control of interacting particles and generalize this to the active control of patterning fields. We demonstrate the applicability of our computational framework via numerical experiments on the control of phase separation with and without a conserved order parameter, self-assembly of fluid droplets, coupled reaction-diffusion equations and finally a phenomenological model for spatio-temporal tissue differentiation. We interpret our numerical experiments in terms of a theoretical understanding of how the underlying physics shapes the geometry of the pattern manifold, altering the transport paths of patterns and the nature of pattern interpolation. We finally conclude by showing how optimal control can be utilized to generate complex patterns via an iterative control protocol over pattern forming pdes which can be casted as gradient flows. All together, our study shows how we can systematically build in physical priors into a generative framework for pattern control in non-equilibrium systems across multiple length and time scales., Comment: 29 pages, 8 figures

PSR J0737$-$3039A/B is unique among double neutron star systems. Its near-perfect edge-on orbit causes the fast spinning pulsar A to be eclipsed by the magnetic field of the slow spinning pulsar B. Using high-sensitivity MeerKAT radio observations combined with updated constraints on the system geometry, we studied the impact of these eclipses on the incident polarization properties of pulsar A. Averaging light curves together after correcting for the rotation of pulsar B revealed enormous amounts of circular polarization and rapid changes in the linear polarization position angle, which occur at phases where emission from pulsar A is partially transmitted through the magnetosphere of pulsar B. These behaviours confirm that the eclipse mechanism is the result of synchrotron absorption in a relativistic pair-plasma confined to the closed-field region of pulsar B's truncated dipolar magnetic field. We demonstrate that changes in circular polarization handedness throughout the eclipses are directly tied to the average line of sight magnetic field direction of pulsar B, from which we unambiguously determine the complete magnetic and viewing geometry of the pulsar., Comment: 8 pages, 6 figures. Accepted for publication in MNRAS

Data rebalancing techniques, including oversampling and undersampling, are a common approach to addressing the challenges of imbalanced data. To tackle unresolved problems related to both oversampling and undersampling, we propose a new undersampling approach that: (i) avoids the pitfalls of noise and overlap caused by synthetic data and (ii) avoids the pitfall of under-fitting caused by random undersampling. Instead of undersampling majority data randomly, our method undersamples datapoints based on their ability to improve model loss. Using improved model loss as a proxy measurement for classification performance, our technique assesses a datapoint's impact on loss and rejects those unable to improve it. In so doing, our approach rejects majority datapoints redundant to datapoints already accepted and, thereby, finds an optimal subset of majority training data for classification. The accept/reject component of our algorithm is motivated by a bilevel optimization problem uniquely formulated to identify the optimal training set we seek. Experimental results show our proposed technique with F1 scores up to 10% higher than state-of-the-art methods.

Logical ring is a core component in membership protocol. However, the logic ring fails to consider the underlying physical latency, resulting in a high diameter. To address this issue, we introduce Diameter-Guided Ring Optimization (DGRO), which focuses on constructing rings with the smallest possible diameter, selecting the most effective ring configurations, and implementing these configurations in parallel. We first explore an integration of deep Q-learning and graph embedding to optimize the ring topology. We next propose a ring selection strategy that assesses the current topology's average latency against a global benchmark, facilitating integration into modern peer-to-peer protocols and substantially reducing network diameter. To further enhance scalability, we propose a parallel strategy that distributes the topology construction process into separate partitions simultaneously. Our experiment shows that: 1) DGRO efficiently constructs a network topology that achieves up to a 60% reduction in diameter compared to the best results from an extensive search over $10^5$ topologies, all within a significantly shorter computation time, 2) the ring selection of DGRO reduces the diameter of state-of-the-art methods Chord, RAPID, and Perigee by 10%-40%, 44%, and 60%. 3) the parallel construction can scale up to $32$ partitions while maintaining the same diameter compared to the centralized version.

The increasing reliance on diffusion models for generating synthetic images has amplified concerns about the unauthorized use of personal data, particularly facial images, in model training. In this paper, we introduce a novel identity inference framework to hold model owners accountable for including individuals' identities in their training data. Our approach moves beyond traditional membership inference attacks by focusing on identity-level inference, providing a new perspective on data privacy violations. Through comprehensive evaluations on two facial image datasets, Labeled Faces in the Wild (LFW) and CelebA, our experiments demonstrate that the proposed membership inference attack surpasses baseline methods, achieving an attack success rate of up to 89% and an AUC-ROC of 0.91, while the identity inference attack attains 92% on LDM models trained on LFW, and the data extraction attack achieves 91.6% accuracy on DDPMs, validating the effectiveness of our approach across diffusion models., Comment: 5 figures, 3 tables,12 pages main body content

In response to challenges in software supply chain security, several organisations have created infrastructures to independently build commodity open source projects and release the resulting binaries. Build platform variability can strengthen security as it facilitates the detection of compromised build environments. Furthermore, by improving the security posture of the build platform and collecting provenance information during the build, the resulting artifacts can be used with greater trust. Such offerings are now available from Google, Oracle and RedHat. The availability of multiple binaries built from the same sources creates new challenges and opportunities, and raises questions such as: 'Does build A confirm the integrity of build B?' or 'Can build A reveal a compromised build B?'. To answer such questions requires a notion of equivalence between binaries. We demonstrate that the obvious approach based on bitwise equality has significant shortcomings in practice, and that there is value in opting for alternative notions. We conceptualise this by introducing levels of equivalence, inspired by clone detection types. We demonstrate the value of these new levels through several experiments. We construct a dataset consisting of Java binaries built from the same sources independently by different providers, resulting in 14,156 pairs of binaries in total. We then compare the compiled class files in those jar files and find that for 3,750 pairs of jars (26.49%) there is at least one such file that is different, also forcing the jar files and their cryptographic hashes to be different. However, based on the new equivalence levels, we can still establish that many of them are practically equivalent. We evaluate several candidate equivalence relations on a semi-synthetic dataset that provides oracles consisting of pairs of binaries that either should be, or must not be equivalent., Comment: 20 pages, 1 figure, 10 tables

We propose a novel optical flow based approach to enhance the axial resolution of anisotropic 3D EM volumes to achieve isotropic 3D reconstruction. Assuming spatial continuity of 3D biological structures in well aligned EM volumes, we reasoned that optical flow estimation techniques, often applied for temporal resolution enhancement in videos, can be utilized. Pixel level motion is estimated between neighboring 2D slices along z, using spatial gradient flow estimates to interpolate and generate new 2D slices resulting in isotropic voxels. We leverage recent state-of-the-art learning methods for video frame interpolation and transfer learning techniques, and demonstrate the success of our approach on publicly available ultrastructure EM volumes.

We study random integer-valued Lipschitz functions on regular trees. It was shown by Peled, Samotij and Yehudayoff that such functions are localized, however, finer questions about the structure of Gibbs measures remain unanswered. Our main result is that the weak limit of a uniformly chosen 1-Lipschitz function with 0 boundary condition on a $d$-ary tree of height $n$ exists as $n \to \infty$ if $2 \le d \le 7$, but not if $d \ge 8$, thereby partially answering a question posed by Peled, Samotij and Yehudayoff. For large $d$, the value at the root alternates between being almost entirely concentrated on 0 for even $n$ and being roughly uniform on $\{-1,0,1\}$ for odd $n$, leading to different limits as $n$ approaches infinity along evens or odds. For $d \ge 8$, the essence of this phenomenon is preserved, which obstructs the convergence. For $d \le 7$, this phenomenon ceases to exist, and the law of the value at the root loses its connection with the parity of $n$. Along the way, we also obtain an alternative proof of localization. The key idea is a fixed point convergence result for a related operator on $\ell^\infty$, and a procedure to show that the iterations get into a `basin of attraction' of the fixed point. We also prove some accompanying analogous `even-odd phenomenon' type results about $M$-lipschitz functions on general non-amenable graphs with high enough expansion (this includes for example the large $d$ case for regular trees). We also prove a convergence result for 1-Lipschitz functions with $\{0,1\}$ boundary condition. This last result relies on an absolute value FKG for uniform 1-Lipschitz functions when shifted by $1/2$., Comment: 36 pages

The Small and Large Magellanic Clouds are the only galaxies outside our own in which radio pulsars have been discovered to date. The sensitivity of the MeerKAT radio interferometer offers an opportunity to search for a population of more distant extragalactic pulsars. The TRAPUM (TRansients And PUlsars with MeerKAT) collaboration has performed a radio-domain search for pulsars and transients in the dwarf star-forming galaxies Sextans A and B, situated at the edge of the local group 1.4 Mpc away. We conducted three 2-hour multi-beam observations at L-band (856-1712 MHz) with the full array of MeerKAT. No pulsars were found down to a radio pseudo-luminosity upper limit of 7.9$\pm$0.4 Jy kpc$^{2}$ at 1400 MHz, which is 28 times more sensitive than the previous limit from the Murriyang telescope. This luminosity is 30 per cent greater than that of the brightest known radio pulsar and sets a cut-off on the luminosity distributions of the entire Sextans A and B galaxies for unobscured radio pulsars beamed in our direction. A Fast Radio Burst was detected in one of the Sextans A observations at a Dispersion Measure (DM) of 737 pc cm$^{-3}$. We believe this is a background event not associated with the dwarf galaxy due to its large DM and its S/N being strongest in the wide-field incoherent beam of MeerKAT., Comment: 11 pages, 9 figures, 5 tables. Accepted for publication in Monthly Notices of the Royal Astronomical Society

The kinesin superfamily of motor proteins is a major driver of anterograde transport of vesicles and organelles within eukaryotic cells via microtubules. Numerous studies have elucidated the step-size, velocities, forces, and navigation ability of kinesins both in reconstituted systems and in live cells. Outside of cells, the kinesin-based transport is physically regulated and can be controlled by obstacles or defects in the path, or the interaction between several motors on the same cargo. To explore the physical control parameters on kinesin-driven transport, we created complex microtubule networks in vitro to test how kinesin cargoes made from quantum dots with one to 10 kinesin motors attached are able to navigate the network. We find that many motors on the quantum dot significantly alter distance walked, time spent bound, the average speed, and the tortuosity of the cargo. We also find that the average mesh size of the microtubule network affects the end-to-end distance of the motion, the run time, average speed and tortuosity of cargoes. Thus, both motor number and network density are physical aspects that regulate where cargoes traverse in space and time., Comment: 7 figures in main text, 1 figure in appendix, 13 tables in appendix

We identify a horizon-skimming limit under which wave equations around large classes of black holes allow a determination of their low-lying (quasi-)degenerate normal modes. Building on our recent work, we use these ``quantum horizon" normal modes to study the thermodynamics of the parent black holes. A key observation is that the UV inputs (the location of the UV regulator, the number of species, and the cut-off in the angular Casimir quantum number) can all be combined into the freedom in a single real parameter. Remarkably, this parameter has an interpretation as the central charge of a holomorphically factorized 2D CFT, and choosing it to be the Kerr-CFT value reproduces the black hole's detailed thermodynamics from the statistical mechanics of normal modes. This perspective provides a heuristic understanding for why the Kerr-CFT central charge is related to the angular momentum of the black hole. The black holes we consider include Kerr-Newman in 3+1 dimensions and Cvetic-Youm in 4+1 dimensions (with all six charges), and they need not be BPS or extremal. Our results show that a refined version of the 't Hooftian quantum gas can be made fully consistent with the thermodynamics of very general black holes. This ``mechanical" approach to the central charge is not directly reliant on asymptotic symmetries in the extremal limit, where the black hole is often unstable., Comment: 38 pages

We present a novel dataset aimed at advancing danger analysis and assessment by addressing the challenge of quantifying danger in video content and identifying how human-like a Large Language Model (LLM) evaluator is for the same. This is achieved by compiling a collection of 100 YouTube videos featuring various events. Each video is annotated by human participants who provided danger ratings on a scale from 0 (no danger to humans) to 10 (life-threatening), with precise timestamps indicating moments of heightened danger. Additionally, we leverage LLMs to independently assess the danger levels in these videos using video summaries. We introduce Mean Squared Error (MSE) scores for multimodal meta-evaluation of the alignment between human and LLM danger assessments. Our dataset not only contributes a new resource for danger assessment in video content but also demonstrates the potential of LLMs in achieving human-like evaluations., Comment: Preprint

AI in Medical Imaging project aims to enhance the National Cancer Institute's (NCI) Image Data Commons (IDC) by developing nnU-Net models and providing AI-assisted segmentations for cancer radiology images. We created high-quality, AI-annotated imaging datasets for 11 IDC collections. These datasets include images from various modalities, such as computed tomography (CT) and magnetic resonance imaging (MRI), covering the lungs, breast, brain, kidneys, prostate, and liver. The nnU-Net models were trained using open-source datasets. A portion of the AI-generated annotations was reviewed and corrected by radiologists. Both the AI and radiologist annotations were encoded in compliance with the the Digital Imaging and Communications in Medicine (DICOM) standard, ensuring seamless integration into the IDC collections. All models, images, and annotations are publicly accessible, facilitating further research and development in cancer imaging. This work supports the advancement of imaging tools and algorithms by providing comprehensive and accurate annotated datasets.

Understanding and modeling the popularity of User Generated Content (UGC) short videos on social media platforms presents a critical challenge with broad implications for content creators and recommendation systems. This study delves deep into the intricacies of predicting engagement for newly published videos with limited user interactions. Surprisingly, our findings reveal that Mean Opinion Scores from previous video quality assessment datasets do not strongly correlate with video engagement levels. To address this, we introduce a substantial dataset comprising 90,000 real-world UGC short videos from Snapchat. Rather than relying on view count, average watch time, or rate of likes, we propose two metrics: normalized average watch percentage (NAWP) and engagement continuation rate (ECR) to describe the engagement levels of short videos. Comprehensive multi-modal features, including visual content, background music, and text data, are investigated to enhance engagement prediction. With the proposed dataset and two key metrics, our method demonstrates its ability to predict engagements of short videos purely from video content., Comment: Accepted to ECCV 2024. Project page: https://github.com/dasongli1/SnapUGC_Engagement

We investigate the dynamics arising out of the propagation of light pulses with different polarizations through a condensate (referred to as a constant background field) with cross coupling described by a coupled nonlinear Schrodinger equation(NLSE) type equation. We then employ Gauge and Darboux transformation approach to bring out the rich dynamics arising out of the background field and cross coupling. The collisional dynamics of bright solitons is found to be inelastic. The constant background field is found to facilitate the periodic localization of light pulses during propagation. We have also unearthed breathers, bright-bright, bright-dark and dark-bright solitons of the coupled NLSE. While the amplitude of breathers oscillate with time as predicted, their maximum(or minimum) amplitude is found to remain a constant and the addition of cross coupling only contributes to the rapid fluctuations in its amplitude over a period of time. In addition, the reinforcement of cross coupling in the presence of constant wave field facilitates the interference of light pulses leading to interesting pattern formation among bright-bright, bright-dark and dark-bright solitons. The highlight of the results is that one obtains various localized excitations like breathers, bright and dark solitons by simply manipulating the amplitude of the constant wave field., Comment: 14 pages, 6 figures, Accepted for Publication in Romanian Reports in Physics (2024)