Search

We theoretically investigate the resonant inelastic X-ray scattering (RIXS) spectra in a quasi-1D chain of weakly coupled frustrated spin-1/2 trimers, as realized in Na$_{2}$Cu$_{3}$Ge$_{4}$O$_{12}$, with Cu $d^{9}$ 1/2 spins. We compute multi-spin correlations contributing to spin-conserving (SC) and spin non-conserving (NSC) RIXS cross-sections using ultra-short core-hole lifetime expansion within the Kramer-Heisenberg formalism. These excitations involve flipping spins of up to three spin-1/2 trimers and include the inelastic neutron scattering (INS) single spin-flip excitations in the lowest order of the NSC channel. We identify the fractionalization of two coupled frustrated trimers in terms of spinons, doublons, and quartons in the spectra evaluated using exact diagonalization, complementing prior studies single spin-spin flip excitation in inelastic neutron scattering. Specifically, we uncover two new high-energy modes at $\omega \approx 2.4J_1$ and $3.0 J_1$ in the NSC and SC channels that are accessible at the Cu $K$-edge and $L$-edge RIXS spectra, which were missing in the INS study. This, therefore, provides pathways to uncover all the possible excitations in coupled trimers. Our work opens new opportunities for understanding the nature of fractionalization and RIXS spectra of frustrated, low-dimensional spin chains.

Low-Rank Adaptation (LoRA) is a popular technique for parameter-efficient fine-tuning of Large Language Models (LLMs). We study how different LoRA modules can be merged to achieve skill composition -- testing the performance of the merged model on a target task that involves combining multiple skills, each skill coming from a single LoRA. This setup is favorable when it is difficult to obtain training data for the target task and when it can be decomposed into multiple skills. First, we identify practically occurring use-cases that can be studied under the realm of skill composition, e.g. solving hard math-word problems with code, creating a bot to answer questions on proprietary manuals or about domain-specialized corpora. Our main contribution is to show that concatenation of LoRAs (CAT), which optimally averages LoRAs that were individually trained on different skills, outperforms existing model- and data- merging techniques; for instance on math-word problems, CAT beats these methods by an average of 43% and 12% respectively. Thus, this paper advocates model merging as an efficient way to solve compositional tasks and underscores CAT as a simple, compute-friendly and effective procedure. To our knowledge, this is the first work demonstrating the superiority of model merging over data mixing for binary skill composition tasks., Comment: 9 pages plus references and appendices

The training data for LLMs embeds societal values, increasing their familiarity with the language's culture. Our analysis found that 44% of the variance in the ability of GPT-4o to reflect the societal values of a country, as measured by the World Values Survey, correlates with the availability of digital resources in that language. Notably, the error rate was more than five times higher for the languages of the lowest resource compared to the languages of the highest resource. For GPT-4-turbo, this correlation rose to 72%, suggesting efforts to improve the familiarity with the non-English language beyond the web-scraped data. Our study developed one of the largest and most robust datasets in this topic area with 21 country-language pairs, each of which contain 94 survey questions verified by native speakers. Our results highlight the link between LLM performance and digital data availability in target languages. Weaker performance in low-resource languages, especially prominent in the Global South, may worsen digital divides. We discuss strategies proposed to address this, including developing multilingual LLMs from the ground up and enhancing fine-tuning on diverse linguistic datasets, as seen in African language initiatives.

We utilize the Sloan Digital Sky Survey (SDSS) extended Baryon Oscillation Spectroscopic Survey (eBOSS) and Baryon Oscillation Spectroscopic Survey (BOSS) catalogs with precise spectroscopic redshifts to estimate the kinematic redshift dipole caused by the proper motion of the Solar system. We find that the velocity extracted from the kinematic dipole is consistent with Cosmic Microwave Background inferred values. Although the small sky coverage and limited number density of the SDSS sources constrain us from obtaining precise and robust measurements, we leverage the redshift dipole method to estimate the kinematic dipole. The velocity measurements in this study are insensitive to intrinsic clustering, associated with the source count dipole. The kinematic dipole measured in this work and its consistency with CMB values do not guarantee isotropy at large scales. The anisotropy (excess dipole) measured with the NRAO VLA Sky Survey (NVSS) and the WISE Catalog (CatWISE) could be due to the intrinsic distribution of galaxies. The results in this work focus solely on the kinematic dipole term., Comment: 13 pages, 4 figures, 4 tables

Identifying conservation laws is central to every subfield of physics, as they illuminate the underlying symmetries and fundamental principles. These laws have far-reaching implications, not only enhancing our theoretical understanding but also enabling practical applications. A prime example can be found in quantum optics: The conservation of orbital angular momentum (OAM) during spontaneous parametric down-conversion (SPDC) enables the generation of a photon pair with entangled OAM. This quantum correlation is commonly attributed to the conservation of the topological charge of a strong coherent pump field driving the process. However, the topological charge of such a classical field only determines the average OAM per photon, so that the total OAM carried by the field has fluctuations as a direct consequence of its photon statistics. In this article, we report on the first realisation of SPDC pumped by a single twisted photon. Our results confirm OAM conservation at the single-photon level and directly transfer to SPDC induced by classical pump fields by averaging over their photon statistics. In addition to verifying a central property of SPDC, our results present the first implementation of cascaded down-conversion in bulk media, setting the stage for experiments on the direct generation of multi-photon high-dimensional entanglement using all degrees of freedom of light.

Measurement-device-independent quantum key distribution (MDI-QKD) was originally proposed as a means to address the issue of detector side-channel attacks and enable finite secure key rates over longer distances. However, the asymmetric characteristics of the channels from the two sources to the measurement device in MDI-QKD impose constraints on successfully extracting a secure key. In this work, we present a plug-and-play scheme for MDI-QKD based on differential phase shift (DPS) encoding. Specifically, we analyze the effects of pulse-width mismatch and polarization mismatch between the pulses arriving at the measurement device. The polarization mismatch is modeled with an assumption of sharing a common reference frame, and the maximum allowable mismatch is found to be 11 degrees. Furthermore, we show that a channel length asymmetry of 176.5 km results in Hong-Ou-Mandel interference visibility of 0.37, thereby leading to zero secure key rates for a polarization-based MDI-QKD protocol. We then present a plug-and-play architecture for DPS-MDI-QKD as a solution to some of these issues, thereby paving the way for practical implementations of MDI protocols.

Autonomous agents powered by large language models (LLMs) have attracted significant research interest. However, the open-source community faces many challenges in developing specialized models for agent tasks, driven by the scarcity of high-quality agent datasets and the absence of standard protocols in this area. We introduce and publicly release xLAM, a series of large action models designed for AI agent tasks. The xLAM series includes five models with both dense and mixture-of-expert architectures, ranging from 1B to 8x22B parameters, trained using a scalable, flexible pipeline that unifies, augments, and synthesizes diverse datasets to enhance AI agents' generalizability and performance across varied environments. Our experimental results demonstrate that xLAM consistently delivers exceptional performance across multiple agent ability benchmarks, notably securing the 1st position on the Berkeley Function-Calling Leaderboard, outperforming GPT-4, Claude-3, and many other models in terms of tool use. By releasing the xLAM series, we aim to advance the performance of open-source LLMs for autonomous AI agents, potentially accelerating progress and democratizing access to high-performance models for agent tasks. Models are available at https://huggingface.co/collections/Salesforce/xlam-models-65f00e2a0a63bbcd1c2dade4, Comment: Technical report for the Salesforce xLAM model series

Multi-layered materials are everywhere, from fiber-reinforced polymer composites (FRPCs) to plywood sheets to layered rocks. When in service, these materials are often exposed to long-term environmental factors, like moisture, temperature, salinity, etc. Moisture, in particular, is known to cause significant degradation of materials like polymers, often resulting in loss of material durability. Hence, it is critical to determine the total diffusion coefficient of multi-layered materials given the coefficients of individual layers. However, the relationship between a multi-layered material's total diffusion coefficient and the individual layers' diffusion coefficients is not well established. Existing parallel and series models to determine the total diffusion coefficient do not account for the order of layer stacking. In this paper, we introduce three parameters influencing the diffusion behavior of multi-layered materials: the ratio of diffusion coefficients of individual layers, the volume fraction of individual layers, and the stacking order of individual layers. Computational models are developed within a finite element method framework to conduct parametric analysis considering the proposed parameters. We propose a new model to calculate the total diffusion coefficient of multi-layered materials more accurately than current models. We verify this parametric study by performing moisture immersion experiments on multi-layered materials. Finally, we propose a methodology for designing and optimizing the cross-section of multi-layered materials considering long-term moisture resistance. This study gives new insights into the diffusion behavior of multi-layered materials, focusing on polymer composites.

Large Transformer networks are increasingly used in settings where low inference latency can improve the end-user experience and enable new applications. However, autoregressive inference is resource intensive and requires parallelism for efficiency. Parallelism introduces collective communication that is both expensive and represents a phase when hardware resources are underutilized. Towards mitigating this, Kraken is an evolution of the standard Transformer architecture that is designed to complement existing tensor parallelism schemes for efficient inference on multi-device systems. By introducing a fixed degree of intra-layer model parallelism, the architecture allows collective operations to be overlapped with compute, decreasing latency and increasing hardware utilization. When trained on OpenWebText, Kraken models reach a similar perplexity as standard Transformers while also preserving their language modeling capabilities when evaluated on the SuperGLUE benchmark. Importantly, when tested on multi-GPU systems using TensorRT-LLM engines, Kraken speeds up Time To First Token by a mean of 35.6% across a range of model sizes, context lengths, and degrees of tensor parallelism.

This paper introduces a novel interface for experiencing music through haptic impulses to the palm of the hand. It presents a practical implementation of the system exploring the realm of musical haptics through the translation of MIDI data from a Digital Audio Workstation (DAW) into haptic sensations, from a set of haptic actuators, in real-time. It also includes a suitable music-to-haptic mapping strategy to translate notes from musical instruments to haptic feedback. The haptic actuators, placed strategically on the palmar surface of the hand allowed users to perceive music and were able to identify melody and rhythm of different musical compositions. A pilot user study conducted intended to assess the accuracy of the interface by testing the participants to select the correct audio presentation from the haptic presentation of the same musical composition. It presents a comparative study, differentiating between those with prior musical background and those without, in identifying the correct audio counterpart solely through haptic inputs. This pilot study delves into how users perceive and interpret haptic feedback within the context of musical compositions. The study showed promising results in enriching our understanding of user responses to haptic feedback in musical scenarios and exploring the intricacies of user experience with the system and its impact on musical interpretation., Comment: 23 pages, 14 figures

We report the experimental observation of fully developed Kolmogorov turbulence originating from self-excited vortex flows in a three-dimensional (3D) dust cloud. The characteristic -5/3 scaling of three-dimensional Kolmogorov turbulence is universally observed in both the spatial and temporal power spectra. Additionally, the 2/3 scaling in the second-order structure function further confirms the presence of Kolmogorov turbulence. We also identified a slight deviation in the tails of the probability distribution functions for velocity gradients. The dust cloud formed in the diffused region away from the electrode and above the glass device surface in the glow discharge experiments. The dust rotation was observed in multiple experimental campaigns under different discharge conditions at different spatial locations and background plasma environments., Comment: 13 pages, 13 figures

Vertical Take-Off and Landing (VTOL) vehicles have gained immense popularity in the delivery drone market and are now being developed for passenger transportation in urban areas to efficiently enable Urban Air Mobility (UAM). UAM aims to utilize the urban airspace \hidetxt{vertical dimension} to address the problem of heavy road congestion in dense urban cities. VTOL vehicles require vertiport terminals for landing, take-off, passengers boarding or deboarding, refuelling (or charging), and maintenance. An efficient scheduling algorithm is essential to maximize the throughput of the vertiport terminal (vertiminal)\hidetxt{ as well as efficient use of airspace} while maintaining safety protocols to handle the UAM traffic. While traditional departure and taxiing operations can be applied in the context of vertiminal, specific algorithms are required for take-off and landing schedules. Unlike fixed-wing aircraft that require a runway to take-off and climb in a single direction, VTOL vehicles can approach and climb in several directions. We propose a Mixed Integer Linear Program (MILP) formulation to schedule flights for taxiing, climbing (or approaching) using multiple directions after take-off (before landing) and turnaround on gates. We also derived equations to thoroughly analyze the throughput capacity of a vertiminal considering all its core elements. We have shown that our MILP can achieve the maximum throughput obtained through the equations. Given the input parameters, our analysis can be used to analyze the capacity of a vertiminal without running any simulation, while our MILP can be used to get the most efficient schedule., Comment: Extension of conference work "Integrated Taxiing and TLOF pad Scheduling Using Different Surface Directions with Fairness Analysis" published in ITSC 2023. DOI: https://doi.org/10.1109/ITSC57777.2023.10422484

While current personal smart devices excel in digital domains, they fall short in assisting users during human environment interaction. This paper proposes Heads Up eXperience (HUX), an AI system designed to bridge this gap, serving as a constant companion across the extended reality (XR) environments. By tracking the user's eye gaze, analyzing the surrounding environment, and interpreting verbal contexts, the system captures and enhances multi-modal data, providing holistic context interpretation and memory storage in real-time task specific situations. This comprehensive approach enables more natural, empathetic and intelligent interactions between the user and HUX AI, paving the path for human computer environment interaction. Intended for deployment in smart glasses and extended reality headsets, HUX AI aims to become a personal and useful AI companion for daily life. By integrating digital assistance with enhanced physical world interactions, this technology has the potential to revolutionize human-AI collaboration in both personal and professional spheres paving the way for the future of personal smart devices., Comment: 48 pages, 16 figures

Non-charge based logic in single-hole spin of semiconductor quantum dots (QDs) can be controlled by anisotropic gate potentials providing a notion for making next generation solid-state quantum devices. In this study, we investigate the isotropic and anisotropic behavior of phonon mediated spin relaxation of heavy-hole spin hot spots in QDs. For the electron spin in isotropic QDs, hot spots are known to be always present due to the Rashba spin-orbit coupling. But for heavy holes in isotropic dots, we show that the occurrences of spin hot spots are sensitive to the bulk g-factor. The hot spot for Rashba coupling in InAs and GaSb dots arises because these materials possess negative bulk g-factor, while that for the Dresselhaus coupling in GaAs and InSb dots is found due to their positive bulk g-factor. For anisotropic QDs, on the other hand, the spin hot spot is universally present due to their broken in-plane rotational symmetry. Further, the increasing electric field, that strengthens the Rashba coupling, is shown to cover a wide range of magnetic field by the hot spots. Results demonstrate that the magnetic field, choice of dot materials and size anisotropy can act as effective control parameters which can be experimentally used to design the device for detecting the phonon mediated heavy-hole spin-relaxation behavior of III-V semiconductor QDs., Comment: 15 pages, 11 figures

One of the most critical tasks of Microsoft sellers is to meticulously track and nurture potential business opportunities through proactive engagement and tailored solutions. Recommender systems play a central role to help sellers achieve their goals. In this paper, we present a content recommendation model which surfaces various types of content (technical documentation, comparison with competitor products, customer success stories etc.) that sellers can share with their customers or use for their own self-learning. The model operates at the opportunity level which is the lowest possible granularity and the most relevant one for sellers. It is based on semantic matching between metadata from the contents and carefully selected attributes of the opportunities. Considering the volume of seller-managed opportunities in organizations such as Microsoft, we show how to perform efficient semantic matching over a very large number of opportunity-content combinations. The main challenge is to ensure that the top-5 relevant contents for each opportunity are recommended out of a total of $\approx 40,000$ published contents. We achieve this target through an extensive comparison of different model architectures and feature selection. Finally, we further examine the quality of the recommendations in a quantitative manner using a combination of human domain experts as well as by using the recently proposed "LLM as a judge" framework.

Blood vessel networks, represented as 3D graphs, help predict disease biomarkers, simulate blood flow, and aid in synthetic image generation, relevant in both clinical and pre-clinical settings. However, generating realistic vessel graphs that correspond to an anatomy of interest is challenging. Previous methods aimed at generating vessel trees mostly in an autoregressive style and could not be applied to vessel graphs with cycles such as capillaries or specific anatomical structures such as the Circle of Willis. Addressing this gap, we introduce the first application of \textit{denoising diffusion models} in 3D vessel graph generation. Our contributions include a novel, two-stage generation method that sequentially denoises node coordinates and edges. We experiment with two real-world vessel datasets, consisting of microscopic capillaries and major cerebral vessels, and demonstrate the generalizability of our method for producing diverse, novel, and anatomically plausible vessel graphs., Comment: Accepted to MICCAI 2024

Models to describe electrical conduction in quantum dot (QD) constituted films often overlook the effects of geometric disorder. We address related issues by examining the influence of the QD packing fraction (PF) on the charge transport and transmission in QD arrays. Using transfer matrix based algorithms and Monte Carlo simulations, we quantify the transmission across disordered QD assemblies. Our results indicate a critical packing fraction ($PF_c $) of $\sim$ 0.64, marking a transition from a non-conducting to a conducting state, aligning well with experimental observations and analytical predictions. This study enhances the understanding of transport in QD arrays, with implications for designing efficient electronic devices based on disordered nanoscale systems., Comment: 5 pages (including references/acknowledgements) and 3 figures

This work presents a soft continuum robot (SCR) that can be used as a soft continuum manipulator (SCM) and a soft snake robot (SSR). This is achieved using expanded polyethylene foam (EPE) modules as the soft material. In situations like post-earthquake search operations, these dual-purpose robots could play a vital role. The soft continuum manipulator with a camera attached to the tip can manually search for survivors in the debris. On the other hand, the soft snake robot can be made by attaching an active wheel to the soft continuum manipulator. This mobile robot can reach places humans cannot and gather information about survivors. This work presents the design, fabrication, and experimental validation of the dual soft continuum robot., Comment: 41 pages, 21 figures

The interaction between boron and hydrogen at grain boundaries has been investigated experimentally and numerically in boron-doped and boron-free martensitic steels using thermal desorption spectrometry (TDS) and ab initio calculations. The calculations show that boron and hydrogen are attracted to grain boundaries but boron can repel hydrogen. This behavior has also been observed using TDS measurements, with the disappearance of one peak when boron is incorporated into the microstructure. Additionally, the microstructure of both steels has been studied through electron backscattered diffraction, electron channeling contrast imaging, synchrotron X-ray measurements, and atom probe tomography. While they have a similar grain size, grain boundary distribution, and dislocation densities, a pronounced boron segregation into PAGBs is observed for boron-doped steels. Then, the equilibrium hydrogen concentration in different trapping sites has been evaluated using the Langmuir-McLean approximation. This thermodynamic model shows that the distribution of hydrogen is identical for all traps when the total hydrogen concentration is low for boron-free steel. However, when it increases, traps of the lowest segregation energies (mostly PAGBs) are firstly saturated, which promotes failure initiation at this defect type. This finding partially explains why PAGBs are the weakest microstructure feature when martensitic steels are exposed to hydrogen-containing environments., Comment: Pre-printed version, also submitted to hal science: https://hal.science/hal-04631250

This paper investigates the impact of weave architectures on the mechanics of crack propagation in fiber-reinforced woven polymer composites under quasi-static loading. Woven composites consist of fabrics/textiles containing fibers interwoven at 0 degrees (warp) and 90 degrees (weft) bound by a polymer matrix. The mechanical properties can be tuned by weaving fiber bundles with single or multiple materials in various patterns or architectures. Although the effects of uniform weave architectures, like plain, twill, satin, etc. on in-plane modulus and fracture energy have been studied, the influence of patterned weaves consisting of a combination of sub-patterns, that is, architected weaves, on these behaviors is not understood. We focus on identifying the mechanisms affecting crack path tortuosity and propagation rate in composites with architected woven textiles containing various sub-patterns, hence, \textit{Ar$\chi$i} {\bf(ar.kee)} \textit{-Textile} Composites. Through compact tension tests, we determine how architected weave patterns compared to uniform weaves influence mode-I fracture energy of woven composites due to interactions of different failure modes. Results show that fracture energy increases at transition regions between sub-patterns in architected weave composites, with more tortuous crack propagation and higher resistance to crack growth than uniform weave composites. We also introduce three geometrical parameters - transition, area, and skewness factors - to characterize sub-patterns and their effects on in-plane fracture energy. This knowledge can be exploited to design and fabricate safer lightweight structures for marine and aerospace sectors with enhanced damage tolerance under extreme loads.

Chain-of-Thought (CoT) prompting has been shown to enhance the multi-step reasoning capabilities of Large Language Models (LLMs). However, debates persist about whether LLMs exhibit abstract generalization or rely on shallow heuristics when given CoT prompts. To understand the factors influencing CoT reasoning we provide a detailed case study of the symbolic reasoning task of decoding shift ciphers, where letters are shifted forward some number of steps in the alphabet. We analyze the pattern of results produced by three LLMs -- GPT-4, Claude 3, and Llama 3.1 -- performing this task using CoT prompting. By focusing on a single relatively simple task, we are able to identify three factors that systematically affect CoT performance: the probability of the task's expected output (probability), what the model has implicitly learned during pre-training (memorization), and the number of intermediate operations involved in reasoning (noisy reasoning). We show that these factors can drastically influence task accuracy across all three LLMs; e.g., when tested with GPT-4, varying the output's probability of occurrence shifts accuracy from 26% to 70%. Overall, we conclude that CoT prompting performance reflects both memorization and a probabilistic version of genuine reasoning. Code and data at this https://github.com/aksh555/deciphering_cot, Comment: EMNLP 2024 Findings; 9 pages plus references and appendices

In this work we study the flux density dependence of the redshift distribution of low-frequency radio sources observed in the LOFAR Two-metre Sky Survey (LoTSS) deep fields and apply it to estimate the clustering length of the large-scale structure of the Universe, examining flux density limited samples (1 mJy, 2 mJy, 4 mJy and 8 mJy) of LoTSS wide field radio sources. We utilise and combine the posterior probability distributions of photometric redshift determinations for LoTSS deep field observations from three different fields (Bo\"otes, Lockman hole and ELAIS-N1, together about $26$ square degrees of sky), which are available for between $91\%$ to $96\%$ of all sources above the studied flux density thresholds and observed in the area covered by multi-frequency data. We estimate uncertainties by a bootstrap method. We apply the inferred redshift distribution on the LoTSS wide area radio sources from the HETDEX field (LoTSS-DR1; about $424$ square degrees) and make use of the Limber approximation and a power-law model of three dimensional clustering to measure the clustering length, $r_0$, for various models of the evolution of clustering. We find that the redshift distributions from all three LoTSS deep fields agree within expected uncertainties. We show that the radio source population probed by LoTSS at flux densities above $1$ mJy has a median redshift of at least $0.9$. At $2$ mJy, we measure the clustering length of LoTSS radio sources to be $r_0 = (10.1\pm 2.6) \ h^{-1}$Mpc in the context of the comoving clustering model. Our findings are in agreement with measurements at higher flux density thresholds at the same frequency and with measurements at higher frequencies in the context of the comoving clustering model.

In [13], K. Kaur, S. Kamada et al. posed a problem of finding a virtual knot, if exists, with an unknotting index (n,m), for any pair of non-negative integers (n,m). In this paper, we address this question by providing infinite families of virtual knots with unknotting index (n,m), for a given pair of non-negative integers (n,m). Additionally, we extend our result for virtual links also.

In the West African country of Ghana, depression is a significant issue affecting a large number of women. Despite its importance, the issue received insufficient attention during the COVID-19 pandemic. In developed countries, mobile phones serve as a convenient medium for accessing health information and providers. However, in Ghana, women's access to mobile phones is limited by cultural, social, and financial constraints, hindering their ability to seek mental health information and support. While some women in deprived areas can afford feature phones, such as the Nokia 3310, the lack of advanced smartphone features further restricts their access to necessary health information. This paper reviews the potential of Unstructured Supplementary Service Data (USSD) technology to address these challenges. Unlike Short Messaging Service (SMS), USSD can facilitate data collection, complex transactions, and provide information access without the need for internet connectivity. This research proposes studying the use of USSD to improve access to mental health resources for resource-deprived women in Ghana., Comment: 15 pages, 1 figure

Multiparametric magnetic resonance imaging (mpMRI) is a key tool for assessing breast cancer progression. Although deep learning has been applied to automate tumor segmentation in breast MRI, the effect of sequence combinations in mpMRI remains under-investigated. This study explores the impact of different combinations of T2-weighted (T2w), dynamic contrast-enhanced MRI (DCE-MRI) and diffusion-weighted imaging (DWI) with apparent diffusion coefficient (ADC) map on breast tumor segmentation using nnU-Net. Evaluated on a multicenter mpMRI dataset, the nnU-Net model using DCE sequences achieved a Dice similarity coefficient (DSC) of 0.69 $\pm$ 0.18 for functional tumor volume (FTV) segmentation. For whole tumor mask (WTM) segmentation, adding the predicted FTV to DWI and ADC map improved the DSC from 0.57 $\pm$ 0.24 to 0.60 $\pm$ 0.21. Adding T2w did not yield significant improvement, which still requires further investigation under a more standardized imaging protocol. This study serves as a foundation for future work on predicting breast cancer treatment response using mpMRI.

Quantum random number generation (QRNG) is a resource that is a necessity in the field of cryptography. However, its certification has been challenging. In this article, we certify randomness with the aid of quantum entanglement in a device independent setting, where we choose two-photon interference for source characterisation. The CHSH inequality violation and quantum state tomography are used as independent checks on the measurement devices. These measures ensure the unpredictability of quantum random number generation. This work can be easily extended to faster randomness expansion protocols., Comment: Comments and suggestions are welcomed

In this paper, we introduce the concept of the warping degree for twisted knots, construct an invariant for them, and utilize it to establish a labeling scheme for these knots, known as ``warping labeling". We have identified that a warping labeling can be extended to twisted virtual braids, enabling the creation of a function that remains invariant under all R-moves except the R2 move. By limiting the labeling set to $\mathbb{Z}_2$, we can develop invariants for twisted virtual braids.

In this paper we introduce a definition of time that emerges in terms of the geometry of the configuration space of a dynamical system. We illustrate this, using the Hamilton-Jacobi equation, in various examples: particle mechanics on a fixed energy surface; non-Abelian gauge theories for compact semi-simple Lie groups where the Gauss law presents new features; and General Relativity in $d+1$ dimensions with $d$ the dimension of space. The discussion in General Relativity is like the non-abelian gauge theory case except for the indefiniteness of the de Witt metric in the Einstein-Hamilton-Jacobi equation, which we discuss in some detail. We illustrate the general formula for the emergent time in various examples including de Sitter spacetime and asymptotically AdS spacetimes., Comment: Minor corrections, references updated

While there has been extensive work on generating physics-based adversarial samples recently, an overlooked class of such samples come from physical failures in the camera. Camera failures can occur as a result of an external physical process, i.e. breakdown of a component due to stress, or an internal component failure. In this work, we develop a simulated physical process for generating broken lens as a class of physics-based adversarial samples. We create a stress-based physical simulation by generating particles constrained in a mesh and apply stress at a random point and at a random angle. We perform stress propagation through the mesh and the end result of the mesh is a corresponding image which simulates the broken lens pattern. We also develop a neural emulator which learns the non-linear mapping between the mesh as a graph and the stress propagation using constrained propagation setup. We can then statistically compare the difference between the generated adversarial samples with real, simulated and emulated adversarial examples using the detection failure rate of the different classes and in between the samples using the Frechet Inception distance. Our goal through this work is to provide a robust physics based process for generating adversarial samples.

Tactile perception is important for robotic systems that interact with the world through touch. Touch is an active sense in which tactile measurements depend on the contact properties of an interaction--e.g., velocity, force, acceleration--as well as properties of the sensor and object under test. These dependencies make training tactile perceptual models challenging. Additionally, the effects of limited sensor life and the near-field nature of tactile sensors preclude the practical collection of exhaustive data sets even for fairly simple objects. Active learning provides a mechanism for focusing on only the most informative aspects of an object during data collection. Here we employ an active learning approach that uses a data-driven model's entropy as an uncertainty measure and explore relative to that entropy conditioned on the sensor state variables. Using a coverage-based ergodic controller, we train perceptual models in near-real time. We demonstrate our approach using a biomimentic sensor, exploring "tactile scenes" composed of shapes, textures, and objects. Each learned representation provides a perceptual sensor model for a particular tactile scene. Models trained on actively collected data outperform their randomly collected counterparts in real-time training tests. Additionally, we find that the resulting network entropy maps can be used to identify high salience portions of a tactile scene., Comment: Published at ICRA 2024, 7 pages, 7 figures

Monolithic large language models (LLMs) like GPT-4 have paved the way for modern generative AI applications. Training, serving, and maintaining monolithic LLMs at scale, however, remains prohibitively expensive and challenging. The disproportionate increase in compute-to-memory ratio of modern AI accelerators have created a memory wall, necessitating new methods to deploy AI. Composition of Experts (CoE) is an alternative modular approach that lowers the cost and complexity of training and serving. However, this approach presents two key challenges when using conventional hardware: (1) without fused operations, smaller models have lower operational intensity, which makes high utilization more challenging to achieve; and (2) hosting a large number of models can be either prohibitively expensive or slow when dynamically switching between them. In this paper, we describe how combining CoE, streaming dataflow, and a three-tier memory system scales the AI memory wall. We describe Samba-CoE, a CoE system with 150 experts and a trillion total parameters. We deploy Samba-CoE on the SambaNova SN40L Reconfigurable Dataflow Unit (RDU) - a commercial dataflow accelerator architecture that has been co-designed for enterprise inference and training applications. The chip introduces a new three-tier memory system with on-chip distributed SRAM, on-package HBM, and off-package DDR DRAM. A dedicated inter-RDU network enables scaling up and out over multiple sockets. We demonstrate speedups ranging from 2x to 13x on various benchmarks running on eight RDU sockets compared with an unfused baseline. We show that for CoE inference deployments, the 8-socket RDU Node reduces machine footprint by up to 19x, speeds up model switching time by 15x to 31x, and achieves an overall speedup of 3.7x over a DGX H100 and 6.6x over a DGX A100.

Ensuring safety is the most important factor in connected and autonomous vehicles, especially in emergency braking situations. As such, assessing the safety benefits of one information topology over other is a necessary step towards evaluating and ensuring safety. In this paper, we compare the safety benefits of a cooperative adaptive cruise control which utilizes information from one predecessor vehicle (CACC) with the one that utilizes information from multiple predecessors (CACC+) for the maintenance of spacing under an emergency braking scenario. A constant time headway policy is employed for maintenance of spacing (that includes a desired standstill spacing distance and a velocity dependent spacing distance) between the vehicles in the platoon. The considered emergency braking scenario consists of braking of the leader vehicle of the platoon at its maximum deceleration and that of the following vehicles to maintain the spacing as per CACC or CACC+. By focusing on the standstill spacing distance and utilizing Monte Carlo simulations, we assess the safety benefits of CACC+ over CACC by utilizing the following safety metrics: (1) probability of collision, (2) expected number of collisions, and (3) severity of collision (defined as the relative velocity of the two vehicles at impact). We present and provide discussion of these results.

In this study, we investigate a one-dimensional (1D) unidirectional phononic waveguide embedded within a three-dimensional (3D) hexagonal close-packed phononic crystal, achieved by the introduction of a screw dislocation. This approach does not rely on the non-trivial topological characteristics of the 3D crystal. We discover that this dislocation induces a pair of helical modes, characterized by their orthogonal $x$- and $y$-directional displacements being out of phase by 90 degrees, which results in a distinctive rotational motion. These helical modes demonstrate directional propagation, tightly linked to the helicity of the screw dislocation. Through considerations of symmetry, we reveal that the emergence of these helical modes is governed by the symmetry of the screw dislocation itself. Our findings not only provide insights into the interplay between dislocation-induced symmetry and wave propagation in phononic systems but also open new avenues for designing directionally selective waveguides without relying on the crystal's topological properties., Comment: 13 pages, 4 figures

Quantum cloud computing is an emerging paradigm of computing that empowers quantum applications and their deployment on quantum computing resources without the need for a specialized environment to host and operate physical quantum computers. This paper reviews recent advances, identifies open problems, and proposes future directions in quantum cloud computing. It discusses the state-of-the-art quantum cloud advances, including the various cloud-based models, platforms, and recently developed technologies and software use cases. Furthermore, it discusses different aspects of the quantum cloud, including resource management, quantum serverless, security, and privacy problems. Finally, the paper examines open problems and proposes the future directions of quantum cloud computing, including potential opportunities and ongoing research in this emerging field.

In this paper, we investigate the selection of time headway to ensure robust string stability in connected and autonomous vehicle platoons in the presence of signal noise in Vehicle-to-Vehicle (V2V) communication. In particular, we consider the effect of noise in communicated vehicle acceleration from the predecessor vehicle to the follower vehicle on the selection of the time headway in predecessor-follower type vehicle platooning with a Constant Time Headway Policy (CTHP). Employing a CTHP based control law for each vehicle that utilizes on-board sensors for measurement of position and velocity of the predecessor vehicle and wireless communication network for obtaining the acceleration of the predecessor vehicle, we investigate how time headway is affected by communicated signal noise. We derive constraints on the CTHP controller gains for predecessor acceleration, velocity error and spacing error and a lower bound on the time headway which will ensure robust string stability of the platoon against signal noise. We provide comparative numerical simulations on an example to illustrate the main result.

In this paper, we investigate the effect of signal delay in communicated information in connected and autonomous vehicles. In particular, we relate this delay's effect on the selection of the time headway in predecessor-follower type vehicle platooning with a constant time headway policy (CTHP). We employ a CTHP control law for each vehicle in the platoon by considering two cases: cooperative adaptive cruise control (CACC) strategy where information from only one predecessor vehicle is employed and CACC+ where information from multiple predecessor vehicles is employed. We investigate how the lower bound on the time headway is affected by signal transmission delay due to wireless communication. We provide a systematic approach to the derivation of the lower bound of the time headway and selection of the appropriate CTHP controller gains for predecessor acceleration, velocity error and spacing error which will ensure robust string stability of the platoon under the presence of signal delay. We corroborate the main result with numerical simulations.

The concepts of twisted knot theory and singular knot theory inspire the introduction of singular twisted knot theory. This study showcases similar findings for singular twisted links, including the Alexander theorem and the Markov theorem derived from knot theory. Moreover, in this paper we define singular twisted virtual braids and their monoid structure. Additionally, we provide both a monoid and a reduced monoid presentation for singular twisted virtual braids.

Fiber-reinforced polymer composite (FRPC) materials are used extensively in various industries, such as aerospace, automobiles, and electronics packaging, due to their remarkable specific strength and desirable properties, such as enhanced durability and corrosion resistance. The evolution of thermal properties in FRPCs is crucial for advancing thermal management systems, optimizing material performance, and enhancing energy efficiency across these diverse sectors. Despite significant research efforts to develop new materials with improved thermal properties and reduced thermal degradation, there is a lack of understanding of the thermal transport phenomena considering the influence of microscale reinforcement morphology in these composites. In the current study, we performed experimental investigations complemented by computations to determine the thermal transport properties and associated phenomena in epoxy and carbon fiber-reinforced epoxy composites. The experimental findings were utilized as input data for numerical analysis to examine the impact of fiber morphology and volume fraction in thermal transport phenomena. Our results revealed that composites incorporating non-circular fibers manifested higher thermal conductivity than traditional circular fibers in the transverse direction. This can be attributed to increased interconnected heat flow pathways facilitated by the increased surface area of non-circular fibers with the same cross-sectional areas, resulting in efficient heat transfer.

In this work, we study ad conversion measurement, a central functionality in digital advertising, where an advertiser seeks to estimate advertiser website (or mobile app) conversions attributed to ad impressions that users have interacted with on various publisher websites (or mobile apps). Using differential privacy (DP), a notion that has gained in popularity due to its strong mathematical guarantees, we develop a formal framework for private ad conversion measurement. In particular, we define the notion of an operationally valid configuration of the attribution rule, DP adjacency relation, contribution bounding scope and enforcement point. We then provide, for the set of configurations that most commonly arises in practice, a complete characterization, which uncovers a delicate interplay between attribution and privacy., Comment: To appear in PoPETS 2024

Neural Radiance Fields (NeRF) accomplishes photo-realistic novel view synthesis by learning the implicit volumetric representation of a scene from multi-view images, which faithfully convey the colorimetric information. However, sensor noises will contaminate low-value pixel signals, and the lossy camera image signal processor will further remove near-zero intensities in extremely dark situations, deteriorating the synthesis performance. Existing approaches reconstruct low-light scenes from raw images but struggle to recover texture and boundary details in dark regions. Additionally, they are unsuitable for high-speed models relying on explicit representations. To address these issues, we present Thermal-NeRF, which takes thermal and visible raw images as inputs, considering the thermal camera is robust to the illumination variation and raw images preserve any possible clues in the dark, to accomplish visible and thermal view synthesis simultaneously. Also, the first multi-view thermal and visible dataset (MVTV) is established to support the research on multimodal NeRF. Thermal-NeRF achieves the best trade-off between detail preservation and noise smoothing and provides better synthesis performance than previous work. Finally, we demonstrate that both modalities are beneficial to each other in 3D reconstruction., Comment: 25 pages, 13 figures

Uncertainty in medical image segmentation tasks, especially inter-rater variability, arising from differences in interpretations and annotations by various experts, presents a significant challenge in achieving consistent and reliable image segmentation. This variability not only reflects the inherent complexity and subjective nature of medical image interpretation but also directly impacts the development and evaluation of automated segmentation algorithms. Accurately modeling and quantifying this variability is essential for enhancing the robustness and clinical applicability of these algorithms. We report the set-up and summarize the benchmark results of the Quantification of Uncertainties in Biomedical Image Quantification Challenge (QUBIQ), which was organized in conjunction with International Conferences on Medical Image Computing and Computer-Assisted Intervention (MICCAI) 2020 and 2021. The challenge focuses on the uncertainty quantification of medical image segmentation which considers the omnipresence of inter-rater variability in imaging datasets. The large collection of images with multi-rater annotations features various modalities such as MRI and CT; various organs such as the brain, prostate, kidney, and pancreas; and different image dimensions 2D-vs-3D. A total of 24 teams submitted different solutions to the problem, combining various baseline models, Bayesian neural networks, and ensemble model techniques. The obtained results indicate the importance of the ensemble models, as well as the need for further research to develop efficient 3D methods for uncertainty quantification methods in 3D segmentation tasks., Comment: initial technical report

In theory, quantum key distribution (QKD) provides unconditional security; however, its practical implementations are susceptible to exploitable vulnerabilities. This investigation tackles the constraints in practical QKD implementations using weak coherent pulses. We improve on the conventional approach of using decoy pulses by integrating it with the coincidence detection (CD) protocol. Additionally, we introduce an easy-to-implement algorithm to compute asymptotic key rates for the protocol. Furthermore, we have carried out an experimental implementation of the protocol, where we demonstrate that monitoring coincidences in the decoy state protocol leads to enhanced key rates under realistic experimental conditions., Comment: 19 pages, 3 figures. Comments are welcome

In the realm of recommender systems, the ubiquitous adoption of deep neural networks has emerged as a dominant paradigm for modeling diverse business objectives. As user bases continue to expand, the necessity of personalization and frequent model updates have assumed paramount significance to ensure the delivery of relevant and refreshed experiences to a diverse array of members. In this work, we introduce an innovative meta-learning solution tailored to the personalization of models for individual members and other entities, coupled with the frequent updates based on the latest user interaction signals. Specifically, we leverage the Model-Agnostic Meta Learning (MAML) algorithm to adapt per-task sub-networks using recent user interaction data. Given the near infeasibility of productionizing original MAML-based models in online recommendation systems, we propose an efficient strategy to operationalize meta-learned sub-networks in production, which involves transforming them into fixed-sized vectors, termed meta embeddings, thereby enabling the seamless deployment of models with hundreds of billions of parameters for online serving. Through extensive experimentation on production data drawn from various applications at LinkedIn, we demonstrate that the proposed solution consistently outperforms the baseline models of those applications, including strong baselines such as using wide-and-deep ID based personalization approach. Our approach has enabled the deployment of a range of highly personalized AI models across diverse LinkedIn applications, leading to substantial improvements in business metrics as well as refreshed experience for our members.

NLP has recently made exciting progress toward training language models (LMs) with strong scientific problem-solving skills. However, model development has not focused on real-life use-cases of LMs for science, including applications in education that require processing long scientific documents. To address this, we introduce TutorEval and TutorChat. TutorEval is a diverse question-answering benchmark consisting of questions about long chapters from STEM textbooks, written by experts. TutorEval helps measure real-life usability of LMs as scientific assistants, and it is the first benchmark combining long contexts, free-form generation, and multi-disciplinary scientific knowledge. Moreover, we show that fine-tuning base models with existing dialogue datasets leads to poor performance on TutorEval. Therefore, we create TutorChat, a dataset of 80,000 long synthetic dialogues about textbooks. We use TutorChat to fine-tune Llemma models with 7B and 34B parameters. These LM tutors specialized in math have a 32K-token context window, and they excel at TutorEval while performing strongly on GSM8K and MATH. Our datasets build on open-source materials, and we release our models, data, and evaluations., Comment: 8 pages without bibliography and appendix, 26 pages total

Let $\theta$ be a root of a monic polynomial $h(x) \in \Z[x]$ of degree $n \geq 2$. We say $h(x)$ is monogenic if it is irreducible over $\Q$ and $\{ 1, \theta, \theta^2, \ldots, \theta^{n-1} \}$ is a basis for the ring $\Z_K$ of integers of $K = \Q(\theta)$. In this article, we study about the monogenity of number fields generated by a root of composition of two binomials. We characterise all the primes dividing the index of the subgroup $\Z[\theta]$ in $\Z_K$ where $K = \Q(\theta)$ with $\theta$ having minimal polynomial $F(x) = (x^m-b)^n - a \in \Z[x]$, $m\geq 1$ and $n \geq 2$. As an application, we provide a class of pairs of binomials $f(x)=x^n-a$ and $g(x)=x^m-b$ having the property that both $f(x)$ and $f(g(x))$ are monogenic., Comment: 9 pages

Remote labs are a groundbreaking development in the education industry, providing students with access to laboratory education anytime, anywhere. However, most remote labs are costly and difficult to scale, especially in developing countries. With this as a motivation, this paper proposes a new remote labs (RLabs) solution that includes two use case experiments: Vanishing Rod and Focal Length. The hardware experiments are built at a low-cost by retrofitting Internet of Things (IoT) components. They are also made portable by designing miniaturised and modular setups. The software architecture designed as part of the solution seamlessly supports the scalability of the experiments, offering compatibility with a wide range of hardware devices and IoT platforms. Additionally, it can live-stream remote experiments without needing dedicated server space for the stream. The software architecture also includes an automation suite that periodically checks the status of the experiments using computer vision (CV). RLabs is qualitatively evaluated against seven non-functional attributes - affordability, portability, scalability, compatibility, maintainability, usability, and universality. Finally, user feedback was collected from a group of students, and the scores indicate a positive response to the students' learning and the platform's usability., Comment: 30 pages, 7 tables and 20 figures. Submitted to ACM Transactions on IoT

We present a physics-based model that accurately predicts the performance of Medtronic's implantable medical device battery lithium/carbon monofluoride (CF$_x$) - silver vanadium oxide (SVO) under both low-rate background monitoring and high-rate pulsing currents. The distinct properties of multiple active materials are reflected by parameterizing their thermodynamics, kinetics, and mass transport properties separately. Diffusion limitations of Li$^+$ in SVO are used to explain cell voltage transient behavior during pulse and post-pulse relaxation. We also introduce change in cathode electronic conductivity, Li metal anode surface morphology, and film resistance buildup to capture evolution of cell internal resistance throughout multi-year electrical tests. We share our insights on how the Li$^+$ redistribution process between active materials can restore pulse capability of the hybrid electrode, allow CF$_x$ to indirectly contribute to capacity release during pulsing, and affect the operation protocols and design principles of batteries with other hybrid electrodes. We also discuss additional complexities in porous electrode model parameterization and electrochemical characterization techniques due to parallel reactions and solid diffusion pathways across active materials. We hope our models implemented in the Hybrid Multiphase Porous Electrode Theory (Hybrid-MPET) framework can complement future experimental research and accelerate development of multi-active material electrodes with targeted performance., Comment: For code and sample usage, please visit: https://github.com/HarryQL/Hybrid-MPET/tree/medtronic_pulse