Search

The allostatic load index (ALI) is a composite measure of whole-person health. Data from electronic health records (EHR) present a huge opportunity to operationalize the ALI in the learning health system, except they are prone to missingness and errors. Validation of EHR data (e.g., through chart reviews) can provide better-quality data, but realistically, only a subset of patients' data can be validated, and most protocols do not recover missing data. Using a representative sample of 1000 patients from the EHR at an extensive learning health system (100 of whom could be validated), we propose methods to design, conduct, and analyze statistically efficient and robust studies of the ALI and healthcare utilization. With semiparametric maximum likelihood estimation, we robustly incorporate all available data into statistical models. Using targeted design strategies, we examine ways to select the most informative patients for validation. Incorporating clinical expertise, we devise a novel validation protocol to promote the quality and completeness of EHR data. Validating the EHR data uncovered relatively low error rates and recovered some missing data. Through simulation studies based on preliminary data, residual sampling was identified as the most informative strategy for completing our validation study. Statistical models of partially validated data indicated higher odds of engaging in the healthcare system were associated with worse whole-person health (i.e., higher ALI), adjusting for age. Targeted validation with an enriched protocol allowed us to ensure the quality and promote the completeness of the EHR. Findings from our validation study were incorporated into analyses as we operationalize the ALI as a whole-person health measure intended to predict healthcare utilization in the academic learning health system., Comment: 16 pages, 2 tables, 6 figures, supplementary materials and code on GitHub

In this paper, we present an enhanced Convolutional Neural Network (CNN)-based rooftop solar photovoltaic (PV) panel detection approach using satellite images. We propose to use pre-trained CNN-based model to extract the local convolutional features of rooftops. These local features are then combined using the Vectors of Locally Aggregated Descriptors (VLAD) technique to obtain rooftop-level global features, which are then used to train traditional Machine Learning (ML) models to identify rooftop images that do and do not contain PV panels. On the dataset used in this study, the proposed approach achieved rooftop-PV classification scores exceeding the predefined threshold of 0.9 across all three cities for each of the feature extractor networks evaluated. Moreover, we propose a 3-phase approach to enable efficient utilization of the previously trained models on a new city or region with limited labelled data. We illustrate the effectiveness of this 3-phase approach for multi-city rooftop-PV detection task., Comment: Accepted at CODS-COMAD 2024, December, 2024, Jodhpur, India (https://cods-comad.in/accepted-papers.php)

We present Atacama Compact Array (ACA) Band-3 observations of the protocluster SPT2349$-$56, an extreme system hosting ${\gtrsim}\,12$ submillimeter galaxies (SMGs) at $z\,{=}\,4.3$, to study its integrated molecular gas content via CO(4-3) and long-wavelength dust continuum. The $\sim$30-hour integration represents one of the longest exposures yet taken on a single pointing with the ACA 7-m. The low-resolution ACA data ($21.0''\,{\times}\,12.2''$) reveal a 75% excess CO(4-3) flux compared to the sum of individual sources detected in higher-resolution Atacama Large Millimeter Array (ALMA) data ($1.0''\,{\times}\,0.8''$). Our work also reveals a similar result by tapering the ALMA data to $10''$. In contrast, the 3.2mm dust continuum shows little discrepancy between ACA and ALMA. A single-dish [CII] spectrum obtained by APEX/FLASH supports the ACA CO(4-3) result, revealing a large excess in [CII] emission relative to ALMA. The missing flux is unlikely due to undetected faint sources but instead suggests that high-resolution ALMA observations might miss extended and low-surface-brightness gas. Such emission could originate from the circum-galactic medium (CGM) or the pre-heated proto-intracluster medium (proto-ICM). If this molecular gas reservoir replenishes the star formation fuel, the overall depletion timescale will exceed 400Myr, reducing the requirement for the simultaneous SMG activity in SPT2349$-$56. Our results highlight the role of an extended gas reservoir in sustaining a high star formation rate (SFR) in SPT2349$-$56, and potentially establishing the ICM during the transition phase to a mature cluster., Comment: 14 pages, 6 figures and 3 tables. Submitted to ApJL. Comments are welcome!

Realizing universal fault-tolerant quantum computation is a key goal in quantum information science. By encoding quantum information into logical qubits utilizing quantum error correcting codes, physical errors can be detected and corrected, enabling substantial reduction in logical error rates. However, the set of logical operations that can be easily implemented on such encoded qubits is often constrained, necessitating the use of special resource states known as 'magic states' to implement universal, classically hard circuits. A key method to prepare high-fidelity magic states is to perform 'distillation', creating them from multiple lower fidelity inputs. Here we present the experimental realization of magic state distillation with logical qubits on a neutral-atom quantum computer. Our approach makes use of a dynamically reconfigurable architecture to encode and perform quantum operations on many logical qubits in parallel. We demonstrate the distillation of magic states encoded in d=3 and d=5 color codes, observing improvements of the logical fidelity of the output magic states compared to the input logical magic states. These experiments demonstrate a key building block of universal fault-tolerant quantum computation, and represent an important step towards large-scale logical quantum processors., Comment: 8+11 pages, 4+4 figures

Co-Channel Interference (CCI) is a fundamental problem in wireless communication networks. It is a well-studied problem in the field. As channels use the same frequency, interference in the radio waves occurs which, in turn, reduces the capacity of the interfered channels. There is a need to use the least number of frequencies as communication networks advance to 5G. In this paper, we present a novel technique to manage interference on channels. We use time division for links of the same frequency and, as a result, we show a significant reduction in the number of frequencies used overall in the network.

We present ALMA observations of the [CI] 492 and 806$\,$GHz fine-structure lines in 25 dusty star-forming galaxies (DSFGs) at $z\,{=}\,4.3$ in the core of the SPT2349$-$56 protocluster. The protocluster galaxies exhibit a median $L^\prime_{[\text{CI}](2-1)}/L^\prime_{[\text{CI}](1-0)}$ ratio of 0.94 with an interquartile range of 0.81-1.24. These ratios are markedly different to those observed in DSFGs in the field (across a comparable redshift and 850$\,\mu$m flux density range), where the median is 0.55 with an interquartile range of 0.50-0.76, and we show that this difference is driven by an excess of [CI](2-1) in the protocluster galaxies for a given 850$\,\mu$m flux density. We estimate gas excitation temperatures of $T_{\rm ex}\,{=}\,59.1^{+8.1}_{-6.8}\,$K for our protocluster sample and $T_{\rm ex}\,{=}\,33.9^{+2.4}_{-2.2}\,$K for the field sample. Our main interpretation of this result is that the protocluster galaxies have had their cold gas driven to their cores via close-by interactions within the dense environment, leading to an overall increase in the average gas density and excitation temperature, and an elevated [CI](2-1) luminosity-to-far-infrared luminosity ratio., Comment: Submitted to ApJL

Smart contracts form the core of Web3 applications. Contracts mediate the transfer of cryptocurrency, making them irresistible targets for hackers. We introduce ASP, a system aimed at easing the construction of provably secure contracts. The Asp system consists of three closely-linked components: a programming language, a defensive compiler, and a proof checker. The language semantics guarantee that Asp contracts are free of commonly exploited vulnerabilities such as arithmetic overflow and reentrancy. The defensive compiler enforces the semantics and translates Asp to Solidity, the most popular contract language. Deductive proofs establish functional correctness and freedom from critical vulnerabilities such as unauthorized access.

Here, we present the outcomes from the second Large Language Model (LLM) Hackathon for Applications in Materials Science and Chemistry, which engaged participants across global hybrid locations, resulting in 34 team submissions. The submissions spanned seven key application areas and demonstrated the diverse utility of LLMs for applications in (1) molecular and material property prediction; (2) molecular and material design; (3) automation and novel interfaces; (4) scientific communication and education; (5) research data management and automation; (6) hypothesis generation and evaluation; and (7) knowledge extraction and reasoning from scientific literature. Each team submission is presented in a summary table with links to the code and as brief papers in the appendix. Beyond team results, we discuss the hackathon event and its hybrid format, which included physical hubs in Toronto, Montreal, San Francisco, Berlin, Lausanne, and Tokyo, alongside a global online hub to enable local and virtual collaboration. Overall, the event highlighted significant improvements in LLM capabilities since the previous year's hackathon, suggesting continued expansion of LLMs for applications in materials science and chemistry research. These outcomes demonstrate the dual utility of LLMs as both multipurpose models for diverse machine learning tasks and platforms for rapid prototyping custom applications in scientific research., Comment: Updating author information, the submission remains largely unchanged. 98 pages total

On-chip coherent laser sources are crucial for the future of photonic integrated circuits, yet progress has been hindered by the complex interplay between material quality, device geometry, and performance metrics. We combine high-throughput characterization, statistical analysis, experimental design, and multi-objective Bayesian optimization to accelerate the design process for low-threshold, high-yield III-V microring lasers with room-temperature operation at communication wavelengths. We demonstrate a 1.6$\times$ reduction in threshold over expert-designed configurations, achieving a 100% lasing yield that emits within the O-band with a median threshold as low as 33$\mu$J cm$^{-2}$ pulse$^{-1}$.

We present a thorough study of the complexity of optical localized modes in two-dimensional disordered photonic crystals. Direct experimental measurements of complexity were made using an interferometric setup that allowed for extraction of phases and, hence, complex-valued wavefunctions. The comparison of experimental and theoretical results allows us to propose a metric for Anderson localization based on the average value and statistical distribution of complexity. Being an alternative to other known criteria of localization, the proposed metric exploits the openness of the disordered medium and provides a quantitative characterization of the degree of localization allowing for determining the localization length., Comment: Main manuscript 6 pages with 6 figures, Supplementary Information 5 pages, 2 figures

Carroll hydrodynamics arises in the $c\to 0$ limit of relativistic hydrodynamics. Instances of its relevance include the Bjorken and Gubser flow models of heavy-ion collisions, where the ultrarelativistic nature of the flow makes the physics effectively Carrollian. In this paper, we explore the structure of hydrodynamics in what can be termed as the Carrollian regime, where instead of keeping only the leading terms in the $c\to 0$ limit of relativistic hydrodynamics, we perform a small-$c$ expansion and retain the subleading terms as well. We do so both for perfect fluids as well as viscous fluids incorporating first order derivative corrections. As apposite applications of the formalism, we utilize the subleading terms to compute modifications to the Bjorken and Gubser flow equations which bring in, in particular, dependence on rapidity., Comment: 31 pages, 1 figure; v2: 35 pages, minor modifications and improvements, several additional references

Adversarial dynamics are intrinsic to the nature of offense and defense in cyberspace, with both attackers and defenders continuously evolving their technologies. Given the wide array of security products available, users often face challenges in selecting the most effective solutions. Furthermore, traditional benchmarks based on single-point attacks are increasingly inadequate, failing to accurately reflect the full range of attacker capabilities and falling short in properly evaluating the effectiveness of defense products. Automated multi-stage attack simulations offer a promising approach to enhance system evaluation efficiency and aid in analyzing the effectiveness of detection systems. However, simulating a full attack chain is complex and requires significant time and expertise from security professionals, facing several challenges, including limited coverage of attack techniques, a high level of required expertise, and a lack of execution detail. In this paper, we model automatic attack simulation as a planning problem. By using the Planning Domain Definition Language (PDDL) to formally describe the attack simulation problem, and combining domain knowledge of both the problem and the domain space, we enable the planning of attack paths through standardized, domain-independent planning algorithms. We explore the potential of Large Language Models (LLMs) to summarize and analyze knowledge from existing attack documentation and reports, facilitating automated attack planning. We introduce Aurora, a system that autonomously simulates full attack chains based on external attack tools and threat intelligence reports.

We study classical closed bosonic strings probing the near-horizon region of a non-extremal black hole and show that this corresponds to understanding string theory in the Carroll regime. This is done by first performing a Carroll expansion and then a near-horizon expansion of a closed relativistic string, subsequently showing that they agree. Concretely, we expand the phase space action in powers of $c^2$, where $c$ is the speed of light, assuming that the target space admits a string Carroll expansion (where two directions are singled out) and show that there exist two different Carroll strings: a magnetic and an electric string. The magnetic string has a Lorentzian worldsheet, whereas the worldsheet of the electric string is Carrollian. The geometry near the horizon of a four-dimensional (4D) Schwarzschild black hole takes the form of a string Carroll expansion (a 2D Rindler space fibred over a 2-sphere). We show that the solution space of relativistic strings near the horizon bifurcates and the two sectors precisely match with the magnetic/electric Carroll strings with an appropriate target space. Magnetic Carroll strings near a black hole shrink to a point on the two-sphere and either follow null geodesics or turn into folded strings on the 2D Rindler spacetime. Electric Carroll strings wrap the two-sphere and follow a massive geodesic in the Rindler space. Finally, we show that 4D non-extremal Kerr and Reissner-Nordstr\"om black holes also admit string Carroll expansions near their outer horizons, indicating that our formulation extends to generic non-extremal black holes., Comment: 42 pages including 4 appendices, 5 figures

As a two-sided marketplace, Airbnb brings together hosts who own listings for rent with prospective guests from around the globe. Results from a guest's search for listings are displayed primarily through two interfaces: (1) as a list of rectangular cards that contain on them the listing image, price, rating, and other details, referred to as list-results (2) as oval pins on a map showing the listing price, called map-results. Both these interfaces, since their inception, have used the same ranking algorithm that orders listings by their booking probabilities and selects the top listings for display. But some of the basic assumptions underlying ranking, built for a world where search results are presented as lists, simply break down for maps. This paper describes how we rebuilt ranking for maps by revising the mathematical foundations of how users interact with search results. Our iterative and experiment-driven approach led us through a path full of twists and turns, ending in a unified theory for the two interfaces. Our journey shows how assumptions taken for granted when designing machine learning algorithms may not apply equally across all user interfaces, and how they can be adapted. The net impact was one of the largest improvements in user experience for Airbnb which we discuss as a series of experimental validations.

Generative artificial intelligence (GenAI) and large language models (LLMs) have simultaneously opened new avenues for enhancing human learning and increased the prevalence of poor-quality information in student response - termed Botpoop. This study introduces Professor Leodar, a custom-built, Singlish-speaking Retrieval Augmented Generation (RAG) chatbot designed to enhance educational while reducing Botpoop. Deployed at Nanyang Technological University, Singapore, Professor Leodar offers a glimpse into the future of AI-assisted learning, offering personalized guidance, 24/7 availability, and contextually relevant information. Through a mixed-methods approach, we examine the impact of Professor Leodar on learning, engagement, and exam preparedness, with 97.1% of participants reporting positive experiences. These findings help define possible roles of AI in education and highlight the potential of custom GenAI chatbots. Our combination of chatbot development, in-class deployment and outcomes study offers a benchmark for GenAI educational tools and is a stepping stone for redefining the interplay between AI and human learning., Comment: 13 pages, 5 figures, SI with Annexes A, B and C upon request

Bounds on the smallest eigenvalue of the neural tangent kernel (NTK) are a key ingredient in the analysis of neural network optimization and memorization. However, existing results require distributional assumptions on the data and are limited to a high-dimensional setting, where the input dimension $d_0$ scales at least logarithmically in the number of samples $n$. In this work we remove both of these requirements and instead provide bounds in terms of a measure of the collinearity of the data: notably these bounds hold with high probability even when $d_0$ is held constant versus $n$. We prove our results through a novel application of the hemisphere transform., Comment: 47 pages

Recent advances in machine learning techniques have made it possible to use high-throughput screening to identify novel materials with specific properties. However, the large number of potential candidates produced by these techniques can make it difficult to select the most promising ones. In this study, we develop the oxidation state probability (OSP) method which evaluates ternary compounds based on the probability (the OSP metric) of each element to adopt the required oxidation states for fulfilling charge neutrality. We compare this model with Roost and the Fourier-transformed crystal properties (FTCP)-based synthesizability score. Among the top 1000 systems with the most database entries in Materials Project (MP), more than 500 systems exhibit an attested compound among the top 3 compositions when ranked by the OSP metric. We find that the OSP method shows promising results for certain classes of ternary systems, especially those containing nonmetals, s-block, or transition metals. When applied to the Cu-In-Te ternary system, an interesting system for thermoelectric applications, the OSP method predicted the synthesizability of CuIn$_3$Te$_5$ without prior knowledge, and we have successfully synthesized CuIn$_3$Te$_5$ in experiment. Our method has the potential to accelerate the discovery of novel compounds by providing a guide for experimentalists to easily select the most synthesizable candidates from an arbitrarily large set of possible chemical compositions., Comment: 11 pages, 3 figures

We have investigated crystalline AlGaAs/GaAs optical coatings with three ultra-stable cavities operating at 4 K, 16 K, 124 K and 297 K. The response of the resonance frequencies of cavities to variations in optical power indicates effects beyond the photo-thermo-optic effect observed in dielectric coatings. These effects are strongly dependent on the intensity of the intracavity light at 1.5~\textmu m. When the rear side of the mirrors is illuminated with external light, we observe a prominent photo-modified birefringence for photon energies above the GaAs bandgap, which points to a possible mechanism relating our observations to the semiconductor properties of the coatings. Separately, we also present a low maintenance evolution of our 124 K silicon cavity system where the liquid nitrogen based cooling system is replaced with closed cycle cooling from a pulse-tube cryo-cooler.

A blockchain facilitates secure and atomic transactions between mutually untrusting parties on that chain. Today, there are multiple blockchains with differing interfaces and security properties. Programming in this multi-blockchain world is hindered by the lack of general and convenient abstractions for cross-chain communication and computation. Current cross-chain communication bridges have varied and low-level interfaces, making it difficult to develop portable applications. Current methods for multi-chain atomic transactions are limited in scope to cryptocurrency swaps. This work addresses these issues. We first define a uniform, high-level interface for communication between chains. Building on this interface, we formulate a protocol that guarantees atomicity for general transactions whose operations may span several chains. We formulate and prove the desired correctness and security properties of these protocols. Our prototype implementation is built using the LayerZero cross-chain bridge. Experience with this implementation shows that the new abstractions considerably simplify the design and implementation of multi-chain transactions. Experimental evaluation with multi-chain swap transactions demonstrates performance comparable to that of custom-built implementations.

The problem of benign overfitting asks whether it is possible for a model to perfectly fit noisy training data and still generalize well. We study benign overfitting in two-layer leaky ReLU networks trained with the hinge loss on a binary classification task. We consider input data that can be decomposed into the sum of a common signal and a random noise component, that lie on subspaces orthogonal to one another. We characterize conditions on the signal to noise ratio (SNR) of the model parameters giving rise to benign versus non-benign (or harmful) overfitting: in particular, if the SNR is high then benign overfitting occurs, conversely if the SNR is low then harmful overfitting occurs. We attribute both benign and non-benign overfitting to an approximate margin maximization property and show that leaky ReLU networks trained on hinge loss with gradient descent (GD) satisfy this property. In contrast to prior work we do not require the training data to be nearly orthogonal. Notably, for input dimension $d$ and training sample size $n$, while results in prior work require $d = \Omega(n^2 \log n)$, here we require only $d = \Omega\left(n\right)$., Comment: 39 pages

In this report, we introduce the Gemini 1.5 family of models, representing the next generation of highly compute-efficient multimodal models capable of recalling and reasoning over fine-grained information from millions of tokens of context, including multiple long documents and hours of video and audio. The family includes two new models: (1) an updated Gemini 1.5 Pro, which exceeds the February version on the great majority of capabilities and benchmarks; (2) Gemini 1.5 Flash, a more lightweight variant designed for efficiency with minimal regression in quality. Gemini 1.5 models achieve near-perfect recall on long-context retrieval tasks across modalities, improve the state-of-the-art in long-document QA, long-video QA and long-context ASR, and match or surpass Gemini 1.0 Ultra's state-of-the-art performance across a broad set of benchmarks. Studying the limits of Gemini 1.5's long-context ability, we find continued improvement in next-token prediction and near-perfect retrieval (>99%) up to at least 10M tokens, a generational leap over existing models such as Claude 3.0 (200k) and GPT-4 Turbo (128k). Finally, we highlight real-world use cases, such as Gemini 1.5 collaborating with professionals on completing their tasks achieving 26 to 75% time savings across 10 different job categories, as well as surprising new capabilities of large language models at the frontier; when given a grammar manual for Kalamang, a language with fewer than 200 speakers worldwide, the model learns to translate English to Kalamang at a similar level to a person who learned from the same content.

Superparamagnetic colloids present interesting assembly dynamics and propulsion in time-varying magnetic fields due to their magnetic relaxation. However, little is known about the mechanisms governing this magnetic relaxation, which is commonly attributed to the interactions and polydispersity of the ferromagnetic nanoparticles distributed within the colloid. We measure this relaxation from the effective potential between colloids subjected to rotating magnetic fields. Remarkably, our results indicate the presence of magnetic relaxation times much longer than what has been reported, which furthers our understanding of the magnetization of colloids in complex magnetic fields., Comment: 6 pages, 4 figures

By strongly driving a transmon-resonator system, the transmon qubit may eventually escape from its cosine-shaped potential. This process is called transmon ionization (TI) and known to be detrimental to the qubit coherence and operation. In this work, we investigate the onset of TI in an irreversible, parametrically-driven, frequency conversion process in a system consisting of a superconducting 3D-cavity coupled to a fixed-frequency transmon qubit. Above a critical pump power we find a sudden increase in the transmon population. Using Renyi entropy, Floquet modes, and Husimi Q functions, we infer that this abrupt change can be attributed to a quantum-to-classical phase transition. Furthermore, in the context of the single-photon detection, we measure a TI-uncorrected detection efficiency of up to 86% and estimate a TI-corrected value of up to 78% by exploiting the irreversible frequency conversion. Our numerical simulations suggest that increasing the detuning between the pump and qubit frequencies and increasing the qubit anharmonicity can suppress the TI impact. Our findings highlight the general importance of the TI process when operating coupled qubit-cavity systems.

Using integrated spectra for two gravitationally lensed galaxies from the JWST TEMPLATES Early Release Science program, we analyze faint auroral lines, which provide direct measurements of the gas-phase chemical abundance. For the brighter galaxy, SGAS1723$+$34 ($z = 1.3293$), we detect the [OIII]$\lambda4363$, [SIII]$\lambda6312$, and [OII]$\lambda\lambda$7320,7330 auroral emission lines, and set an upper limit for the [NII]$\lambda5755$ line. For the second galaxy, SGAS1226$+$21 ($z = 2.925$), we do not detect any auroral lines, and report upper limits. With these measurements and upper limits, we constrain the electron temperatures in different ionization zones within both of these galaxies. For SGAS1723$+$34, where auroral lines are detected, we calculate direct oxygen and nitrogen abundances, finding an N/O ratio consistent with observations of nearby ($z\sim 0$) galaxies. These observations highlight the potent combination of JWST and gravitational lensing to measure faint emission lines in individual distant galaxies and to directly study the chemical abundance patterns in those galaxies., Comment: Accepted to the Astrophysical Journal's Focus Issue on the TEMPLATES JWST Early Release Science Program. 16 pages, 6 figures