Search

State legislators recognized that students in the United States continue to struggle with reading and have begun implementing laws requiring schools to use science of reading-aligned curriculum when teaching reading (Schwartz, 2022). Many students lack the foundational skills to even begin learning how to read in grades K-3 (Kuhfeld et al., 2022). This study evaluates the effectiveness of Heggerty's Bridge to Reading foundational skills curriculum for first-grade students in a rural, diverse southern school district during the 2023-2024 school year. Using a quasi-experimental design, LXD Research compared reading outcomes of first graders using Bridge to Reading (239 students) with a similar comparison group (724 students). The NWEA MAP Growth assessment showed significant reading gains and higher end-of-year scores for Bridge to Reading students relative to the comparison group. The statistical results were supported by positive feedback from educators and administrators implementing the program through interviews and surveys. These findings demonstrate Bridge to Reading's ability to improve reading achievement for first graders.

This study addresses the important yet underexplored question of whether the Common Core State Standards in Mathematics, which emphasize critical thinking and problem-solving, as well as the computer-based assessments aligned with the Common Core, have facilitated or hindered learning for students with disabilities. By analyzing administrative data from a large county in California, we track mathematics achievement trends before and after the implementation of the Common Core. Our findings show a significant widening of the achievement disparity in mathematics between students with and without disabilities, suggesting that the Common Core and computerized assessments disproportionately affected students with disabilities.

This PACER brief provides RFA's updated analysis of data on students experiencing homelessness in Pennsylvania, highlighting prevalence, challenges with identification, disparities in educational opportunities, and impact on academic outcomes. Our main findings include that: (1) Roughly 2% of students in Pennsylvania were identified as experiencing homelessness in 2021-22, a slight increase from prior years; (2) Pennsylvania has seen a statewide increase in the identification rate of homelessness (the rate of student homelessness/the rate of students in poverty), but continues to rank below the national average, now 30th out of the 50 states; (3) Students experiencing homelessness are heavily concentrated in Pennsylvania's most inadequately funded school districts, which provide fewer educational opportunities, and are increasingly enrolled in cyber charter schools where research has documented consistently poor overall student outcomes; and (4) Pennsylvania's students experiencing homelessness are exhibiting worse academic outcomes compared with all students and other economically disadvantaged students. These findings underscore the urgent need for school leaders and policymakers in Pennsylvania to prioritize new resources and comprehensive strategies to support students experiencing homelessness.

A major open issue concerning the active Sun is the effectiveness with which magnetic reconnection accelerates electrons in flares. A paper published by {\em{Nature}} in 2022 used microwave observations to conclude that the Sun is an almost ideal accelerator, energizing nearly all electrons within a coronal volume to nonthermal energies. Shortly thereafter, a paper published in {\em{Astrophysical Journal Letters}} used hard X-ray measurements \emph{of the same event} to reach the contradictory conclusion that less than 1\% of the available electrons were accelerated. Here we address this controversy by using spatially resolved observations of hard X-ray emission and a spectral inversion method to determine the evolution of the electron spectrum throughout the flare. So we estimated the density of the medium where electrons accelerate and, from this, the ratio of accelerated to ambient electron densities. Results show that this ratio never exceeds a percent or so in the cases analyzed.

Three liquid crystalline equimolar mixtures are formulated, each containing the MHPOBC compound and the homolog from the 3FmHPhF6 series with m = 4, 5 or 6. The properties of the mixtures are investigated by differential scanning calorimetry, polarizing optical microscopy, X-ray diffraction, electro-optic measurements, and UV-Vis-NIR and dielectric spectroscopies. In all mixtures, the smectic A*, smectic C*, and smectic C$_A$* phases are observed. Each mixture can be supercooled and forms the glass of smectic C$_A$* at 225-238 K. The glass transition temperature decreases with the increasing C$_m$H$_{2m}$ chain length in the 3FmHPhF6 molecule. The tilt angle in the smectic C$_A$* phase reaches 36.5-37.5{\deg}. The selective reflection of light in the visible and near-infrared spectral ranges is observed, with a strong dependence of the reflected light's wavenumber on temperature.

Benchmarks are important tools for tracking the rapid advancements in large language model (LLM) capabilities. However, benchmarks are not keeping pace in difficulty: LLMs now achieve over 90\% accuracy on popular benchmarks like MMLU, limiting informed measurement of state-of-the-art LLM capabilities. In response, we introduce Humanity's Last Exam (HLE), a multi-modal benchmark at the frontier of human knowledge, designed to be the final closed-ended academic benchmark of its kind with broad subject coverage. HLE consists of 3,000 questions across dozens of subjects, including mathematics, humanities, and the natural sciences. HLE is developed globally by subject-matter experts and consists of multiple-choice and short-answer questions suitable for automated grading. Each question has a known solution that is unambiguous and easily verifiable, but cannot be quickly answered via internet retrieval. State-of-the-art LLMs demonstrate low accuracy and calibration on HLE, highlighting a significant gap between current LLM capabilities and the expert human frontier on closed-ended academic questions. To inform research and policymaking upon a clear understanding of model capabilities, we publicly release HLE at https://lastexam.ai., Comment: 26 pages, 6 figures

We present an analysis of polarized X-ray pulses based on simulated data for accreting millisecond pulsars (AMPs). We used the open-source X-ray Pulse Simulation and Inference code (previously applied to NICER observations), that we upgraded to allow polarization analysis. We predicted neutron star parameter constraints for the Imaging X-ray Polarimetry Explorer (IXPE) and found that strong limits on the hot region geometries can be hard to obtain if the emitting hot region is large and the number of polarized photons relatively small. However, if the star is bright enough and the hot regions are small and located so that polarization degree is higher, the observer inclination and hotspot colatitude can be constrained to a precision of within a few degrees. We also found that the shape of the hot region, whether a circle or a ring, cannot be distinguished in our most optimistic scenario. Nevertheless, future X-ray polarization missions are expected to improve the constraints, and already the recent AMP polarization detections by IXPE should help to infer the neutron star mass and radius when combined with modelling of X-ray pulse data sets that do not contain polarization information., Comment: 8 pages, 3 figures, 1 table, submitted to MNRAS

We present the spectral analysis of J-138717, a post-starburst galaxy observed at $z=1.8845$ with JWST, for which we estimate a stellar mass M$_* = 3.5\pm0.2\times10^{10}$M$_\odot$ and velocity dispersion $\sigma_* = 198\pm10$ km/s. From the fits of the NIRspec low-resolution prism and medium resolution grating spectra, we estimate an age of $\sim0.9$ Gyr, a metallicity of $\sim -0.3$ dex, and a dust attenuation A$_{\rm V} \sim 0.7$ mag. From the H$\alpha$ and [OII] emission lines we derive a star-formation rate (SFR)$\sim 0.1$ M$_\odot$ yr$^{-1}$, corresponding to a specific SFR$\sim 4 \times 10^{-12}$ yr$^{-1}$. By reconstructing its SFH, we find that this galaxy formed most of its stellar mass quickly ($0.4-0.8$ Gyr) and then rapidly quenched about 0.4 Gyr lookback time. The analysis of the emission lines ([OII], [NeIII], [OIII], H$\alpha$, and [NII]) indicates a low-level residual star formation or a weak AGN as the ionization source. Further, the lack of evidence for ongoing outflows in both the absorption and emission lines tracing the neutral and ionized gas argues against the presence of strong AGN feedback. Thanks to the high quality of the data, we carried out a comprehensive analysis of the NIR spectral indices. This is the first time NIR spectral indices have been investigated at such a high redshift. We find that the observed CO and CN indices are generally weak and in better agreement with models with a lower contribution from thermally pulsating (TP-)AGB stars, indicating that either TP-AGB stars do not contribute significantly to the spectrum of J-138717 or the prescriptions in the models for these evolved phases should be substantially improved. In general, we find that all models struggle to reproduce the NIR indices consistently, pointing to the importance of improving state-of-the-art stellar population models in the NIR, especially at young ages., Comment: submitted to Astronomy & Astrophysics

We present multi-wavelength analysis of ZTF23abelseb (AT 2023sva), an optically discovered fast-fading ($\Delta m_r = 2.2$ mag in $\Delta t = 0.74 $ days), luminous ($M_r \sim -30.0$ mag) and red ($g-r = 0.50$ mag) transient at $z = 2.28$ with accompanying luminous radio emission. AT 2023sva does not possess a $\gamma$-ray burst (GRB) counterpart to an isotropic equivalent energy limit of $E_{\rm{\gamma, \, iso}} < 1.6 \times 10^{52}$ erg, determined through searching $\gamma$-ray satellite archives between the last non-detection and first detection, making it the sixth example of an optically-discovered afterglow with a redshift measurement and no detected GRB counterpart. We analyze AT 2023sva's optical, radio, and X-ray observations to characterize the source. From radio analyses, we find the clear presence of strong interstellar scintillation (ISS) 72 days after the initial explosion, allowing us to place constraints on the source's angular size and bulk Lorentz factor. When comparing the source sizes derived from ISS of orphan events to those of the classical GRB population, we find orphan events have statistically smaller source sizes. We also utilize Bayesian techniques to model the multi-wavelength afterglow. Within this framework, we find evidence that AT 2023sva possesses a shallow power-law structured jet viewed slightly off-axis ($\theta_{\rm{v}} = 0.07 \pm 0.02$) just outside of the jet's core opening angle ($\theta_{\rm{c}} = 0.06 \pm 0.02$). We determine this is likely the reason for the lack of a detected GRB counterpart, but also investigate other scenarios. AT 2023sva's evidence for possessing a structured jet stresses the importance of broadening orphan afterglow search strategies to a diverse range of GRB jet angular energy profiles, to maximize the return of future optical surveys., Comment: 22 pages, 14 Figures, Submitted to MNRAS

JWST observations have unveiled faint active galactic nuclei (AGN) at high-redshift that provide insights on the formation of supermassive black holes (SMBHs) and their coevolution with galaxies. However, disentangling stellar from AGN light in these sources is challenging. Here, we use an empirical approach to infer the average stellar mass of 6 faint broad line (BL) Halpha emitters at z = 4 - 5 with BH masses ~ 6 (4 - 15)x10^6 Msun, with a method independent of their spectral energy distribution (SED). We use the deep JWST/NIRcam grism survey ALT to measure the over-densities around BL-Halpha emitters and around a spectroscopic reference sample of ~300 galaxies. In our reference sample, we find that Mpc-scale over-density correlates with stellar mass, while pair counts are flat below ~50 kpc due to satellites. Their large-scale environments suggest that BL-Halpha emitters are hosted by galaxies with stellar masses ~5x10^7 Msun, ~40 times lower than those inferred from galaxy-only SED fits. Adding measurements around more luminous z~6 AGNs, we find tentative correlations between line width, BH mass and the over-density, suggestive of a steep BH to halo mass relation. The main implications are (1) when BH masses are taken at face value, we confirm extremely high BH to stellar mass ratios of ~10 %, (2) the low stellar mass galaxies hosting growing SMBHs are in tension with typical hydrodynamical simulations, except those without feedback, (3) a 1 % duty cycle implied by the host mass hints at super-Eddington accretion, which may imply over-estimated SMBH masses, (4) the masses are at odds with a high stellar density interpretation of the line broadening, (5) our results imply a diversity of galaxy masses, environments and SEDs among AGN samples, depending on their luminosity., Comment: 19 pages, 15 figures. Submitted to ApJ. Main results in Figures 12, 13, 14. Comments welcome

Food insecurity is widespread among college students in the United States. Food benefits delivered through the CalFresh program, California's version of the federal Supplemental Nutrition Assistance Program (SNAP), can reduce hunger by helping students pay for groceries, but may not reach all eligible students. To date, higher education systems have lacked good estimates of the share of their students who are eligible for CalFresh and the share who actually receive benefits. To address this information gap, the California Policy Lab (CPL) partnered with the California Community College (CCC) Chancellor's Office, the University of California Office of the President (UCOP), the California Department of Social Services (CDSS), and the California Student Aid Commission (CSAC) to build a linked database of student-level administrative data on college enrollment, financial aid, and CalFresh participation. This database covers all students enrolled at CCC or UC campuses from academic years 2010-11 through 2021-22, along with corresponding FAFSA submissions and CalFresh participation. Using these data, researchers are able to measure how many college students are likely eligible for CalFresh, and of those how many participate. This report was updated in August 2024 with updated eligibility statistics that reflect results from a refined simulation of student eligibility for CalFresh. Using the improved simulation, the researchers found that among California Community College students, 20% of students were eligible for CalFresh (an increase from the 16% estimate in the June version of this report), among University of California (UC) undergraduates, 33% were eligible (an increase from 31% in the June report), and among UC graduate students, 7% were eligible (an increase from 6%).

The COVID-19 pandemic devastated student achievement, with declines rivaling those after Hurricane Katrina. These losses widened achievement gaps between historically marginalized students and their peers. Three years later, achievement remains behind pre-pandemic levels for many students. This paper examines 2022-23 academic recovery efforts across eight districts, including tutoring, small group instruction, after-school, extended year, double-dose, digital learning, and expert teacher interventions. Across 22 math and reading interventions, most were delivered to fewer students and for less time than planned. We find positive effects for one tutoring program on math scores and two tutoring programs on reading scores, ranging from 0.22 to 0.33 SD. Each of these programs served a very small share of the district's students and was unlikely to play a major role in district-wide academic recovery. Finally, we find that having an "expert" teacher with high evaluation scores as opposed to a non-expert teacher significantly improves student achievement by 0.06 SD in math and 0.11 SD in reading. While highlighting the promise of intensive academic interventions, our findings underscore the challenges districts face in scaling such interventions to match their recovery needs. The field needs better evidence regarding successful implementation of large-scale interventions.

This report explores the dramatic increase in Head Start teacher turnover during the COVID-19 pandemic. Using data from the Head Start Program Information Report (PIR) from 2010 to 2022, the study investigates trends in teacher turnover across all Head Start program types and examines the reasons behind teacher departures. The report finds that teacher turnover surged significantly in both 2020-21 and 2021-22, with the 2021-22 year experiencing a particularly sharp increase exceeding 28% of teachers. Turnover rates varied by program type, with Early Head Start programs exhibiting higher turnover than traditional Head Start programs, and American Indian Alaska Native (AIAN) Early Head Start programs seeing the highest rate at 37.1%. The study reveals that compensation remains the most common reason cited for teachers' departures, both before and during the pandemic. However, COVID-19 related factors also emerged as prominent reasons for departure. This research underscores the urgent need to address the challenges of teacher turnover in the Head Start sector, particularly in the wake of the pandemic. The report advocates for policy initiatives aimed at improving teacher retention through enhanced compensation.

Global progress in child labour reduction has stalled since 2016. In South Asia, the negative consequences of the COVID-19 pandemic, recurrent extreme weather events and the cost-of-living crisis further threaten to slow or even reverse the progress made. This synthesis report addresses a timely need for new research to help decision-makers and practitioners identify policy and programmatic actions that can support prevention and reduction of child labour in South Asia, with a focus on the role of education. Drawing on quantitative data and qualitative primary evidence from UNICEF Innocenti's research in Bangladesh and India, it offers an accessible and rigorous overview of recent trends and patterns in child labour and explores how these relate to children's schooling experiences in both countries. Implications of COVID-19 for education and labour, and the role of child marriage as a key related outcome, are also assessed. The report also explores children's perspectives of their labour and schooling experiences, including in migration contexts as well as during the COVID-19 pandemic and resulting school closures. The report concludes with programmatic and research recommendations for further action at regional, country, and global levels.

A five-dimensional Kaluza-Klein spacetime model is considered, with one extra compactified spatial dimension. The equation of state of an electrically neutral, zero-temperature Fermi gas with a repulsive linear potential is described. From the equation of state, the speed of sound squared is calculated and shown for different model parameters. Its properties are studied from lower energies up to the conformal limit., Comment: 7 pages, 1 figure

The World Wide Web is a complex interconnected digital ecosystem, where information and attention flow between platforms and communities throughout the globe. These interactions co-construct how we understand the world, reflecting and shaping public discourse. Unfortunately, researchers often struggle to understand how information circulates and evolves across the web because platform-specific data is often siloed and restricted by linguistic barriers. To address this gap, we present a comprehensive, multilingual dataset capturing all Wikipedia links shared in posts and comments on Reddit from 2020 to 2023, excluding those from private and NSFW subreddits. Each linked Wikipedia article is enriched with revision history, page view data, article ID, redirects, and Wikidata identifiers. Through a research agreement with Reddit, our dataset ensures user privacy while providing a query and ID mechanism that integrates with the Reddit and Wikipedia APIs. This enables extended analyses for researchers studying how information flows across platforms. For example, Reddit discussions use Wikipedia for deliberation and fact-checking which subsequently influences Wikipedia content, by driving traffic to articles or inspiring edits. By analyzing the relationship between information shared and discussed on these platforms, our dataset provides a foundation for examining the interplay between social media discourse and collaborative knowledge consumption and production.

We present a theoretical study of the pump-probe Faraday rotation and ellipticity signals in ensembles of uniaxially anisotropic CdSe nanoplatelets and nanocrystals. We use the Faraday rotation mechanism based on the excitation of negative heavy hole trions for a magnetic field applied in the Voigt geometry. Three types of ensembles with typical spatial distributions of the orientation of the anisotropy axis with respect to the direction of light propagation are considered. Faraday rotation and ellipticity signals are modeled for excitation by single and repeated pump pulses, taking into account the anisotropy of the electron g-factor. We show that spin dephasing caused by the electron g-factor anisotropy and the arbitrary orientation of nanoplatelets or nanocrystals result only in partial damping of oscillation amplitude in contrast to the dephasing caused by the dispersion of the electron g-factor in the ensemble. We demonstrate that regardless of the g-factor anisotropy degree the oscillation frequency of the Faraday rotation and ellipticity signals for a randomly oriented ensemble is determined by the transverse electron g-factor component., Comment: 13 pages, 6 figures

Agency is a system's capacity to steer outcomes toward a goal, and is a central topic of study across biology, philosophy, cognitive science, and artificial intelligence. Determining if a system exhibits agency is a notoriously difficult question: Dennett (1989), for instance, highlights the puzzle of determining which principles can decide whether a rock, a thermostat, or a robot each possess agency. We here address this puzzle from the viewpoint of reinforcement learning by arguing that agency is fundamentally frame-dependent: Any measurement of a system's agency must be made relative to a reference frame. We support this claim by presenting a philosophical argument that each of the essential properties of agency proposed by Barandiaran et al. (2009) and Moreno (2018) are themselves frame-dependent. We conclude that any basic science of agency requires frame-dependence, and discuss the implications of this claim for reinforcement learning.

Accurate calculations of the nucleation rate $\Gamma$ for first order phase transitions are important for determining their observable consequences in particle physics and cosmology. Perturbative calculations are often used, but they are incomplete and should be tested against fully nonperturbative lattice simulations. We simulate nucleation on the lattice in a scalar field theory with a tree-level barrier, a scenario which should be well described by perturbation theory. Our computation of the nucleation rate, however, only shows qualitative agreement with the perturbative result. This motivates further study of nucleation on the lattice and to higher orders in perturbation theory., Comment: 10 pages, 4 figures, Contribution to The 41st International Symposium on Lattice Field Theory, 28 July - 3 August 2024

Following in the footsteps of the LHC, the Future Circular Collider (FCC) plans to be the next multi-generational collider project. In the first stage, FCC-ee will collide intense beams of electrons and positrons at centre of mass energies between 88 and 365 GeV, making it an electroweak, flavour, Higgs and top factory. The unprecedented statistical precision requires FCC-ee experiments to limit their systematic uncertainties to the very minimum. The precise reconstruction of the interaction vertices is central to most measurements at FCC-ee, such as rare flavour physics processes and the measurement of Higgs and Z decays to bottom and charm quarks and taus. This contribution will discuss the requirements of FCC-ee vertex detectors, from the necessary impact parameter resolution via the challenging collision environment at the Z pole to the tight requirement on the material budget, which should be kept below 0.3% of a radiation length per detection layer. Next, the proposed vertex detector designs for FCC-ee are shortly presented, and an outlook is given on novel detector designs and features. The requirements for the vertexing performance translate into requirements for the sensors used for the vertex detector. As discussed in this contribution, they need to feature a spatial resolution of about 3 $\mu$m and provide timing information of O($\mu$s-ns) while keeping power consumption minimal to allow for air-cooling of the detector - minimising the detector material budget. The only type of sensor capable of aiming to fulfil such requirements are CMOS Monolithic Active Pixel Sensors (MAPS), which combine signal generation, amplification and readout into a single silicon die. Therefore, the rest of this contribution will present an overview of existing and planned MAPS technologies and prototypes aiming to fulfil the stringent FCC-ee vertex detector requirements., Comment: 6 pages, 4 figures, 3 tables, Pixel 2024 conference proceeding

Large Language Models (LLMs) have attained human-level accuracy on medical question-answer (QA) benchmarks. However, their limitations in navigating open-ended clinical scenarios have recently been shown, raising concerns about the robustness and generalizability of LLM reasoning across diverse, real-world medical tasks. To probe potential LLM failure modes in clinical problem-solving, we present the medical abstraction and reasoning corpus (M-ARC). M-ARC assesses clinical reasoning through scenarios designed to exploit the Einstellung effect -- the fixation of thought arising from prior experience, targeting LLM inductive biases toward inflexible pattern matching from their training data rather than engaging in flexible reasoning. We find that LLMs, including current state-of-the-art o1 and Gemini models, perform poorly compared to physicians on M-ARC, often demonstrating lack of commonsense medical reasoning and a propensity to hallucinate. In addition, uncertainty estimation analyses indicate that LLMs exhibit overconfidence in their answers, despite their limited accuracy. The failure modes revealed by M-ARC in LLM medical reasoning underscore the need to exercise caution when deploying these models in clinical settings., Comment: 14 pages, 6 figures

The mutation process in evolution strategies has been interlinked with the normal distribution since its inception. Many lines of reasoning have been given for this strong dependency, ranging from maximum entropy arguments to the need for isotropy. However, some theoretical results suggest that other distributions might lead to similar local convergence properties. This paper empirically shows that a wide range of evolutionary strategies, from the (1+1)-ES to CMA-ES, show comparable optimization performance when using a mutation distribution other than the standard Gaussian. Replacing it with, e.g., uniformly distributed mutations, does not deteriorate the performance of ES, when using the default adaptation mechanism for the strategy parameters. We observe that these results hold not only for the sphere model but also for a wider range of benchmark problems.

We present the stellar properties of 2908 galaxies at 0.6 < z < 1.0 from the LEGA-C survey. We emphasize the importance of high signal-to-noise, high spectral resolution spectroscopy in the inference of stellar population properties of galaxies. We estimate the galaxy properties with the SED fitting code Prospector, by fitting spectroscopy and broadband photometry together, drawn from the LEGA-C DR3 and UltraVISTA catalogs respectively. We report a positive correlation between light-weighted ages and stellar velocity dispersion ($\sigma_\star$). The trend with $\sigma_\star$ is weaker for the mass-weighted ages and stellar metallicity ($Z_\star$). On average, quiescent galaxies are characterized by high $Z_\star$, they are \sim 1.1 Gyr older, less dusty, with steeper dust attenuation slopes compared to star-forming galaxies. Conversely, star-forming galaxies are characterized by significantly higher dust optical depths and shallower (grayer) attenuation slopes. Low mass (high mass) star-forming galaxies have lower (higher) $Z_\star$, while their stellar populations are on average younger (older). A key pragmatic result of our study is that a linear-space metallicity prior is preferable to a logarithmic-space one when using photometry alone, as the latter biases the posteriors downward. Spectroscopy greatly improves stellar population measurements and is required to provide meaningful constraints on age, metallicity, and other properties. Pairing spectroscopy with photometry helps resolving the dust-age-metallicity degeneracy, yielding more accurate mass- and light-weighted ages, with ages inferred from photometry alone suffering such large uncertainties. Stellar metallicities are constrained by our spectroscopy, but precise measurements remain challenging (and impossible with photometry alone), particularly in the absence of Mg and Fe lines redward of 5000 $\AA$ in the observed spectrum., Comment: Accepted, 25 pages, 17 figures

The aim of our paper is twofold. First, we present applied mathematics results on the heat content of bounded domains with smooth boundary and with fractal boundary. Second, we discuss new findings and concepts related to fractal curvatures for domains with fractal boundary. We conjecture that fractal curvatures and their scaling exponents will emerge in the short-time heat content asymptotics of domains with fractal boundary.

Drone-based beam measurements are a promising avenue to tackle the critical challenge of calibration for 21 cm cosmology telescopes. In this paper, we introduce a new drone-based calibration system for 400-800 MHz radio observatories, describing its instrumentation and first deployment. We discuss measurements of the TONE array, a CHIME/FRB outrigger pathfinder, and present results, including full 2D high spatial resolution beam maps in both co- and cross-polarization, as well as comparisons to simulations. The polarized beam maps cover a 70 degree by 70 degree grid, capturing the first two sidelobes and measuring the TONE main beam and first sidelobe with 7-9% statistical errors. We investigate polarization angle alignment with frequency, finding significant polarization leakage in the TONE antennas at frequencies above 600 MHz, and a polarization axis rotation with frequency. We describe statistical and systematic errors, as well as measurements of radio frequency interference from the drone and equipment. Our drone system is the first to incorporate a broad-band switched calibration source in the drone payload, enabling background subtraction and direct measurements of the RFI emitted by the drone. The results presented are the first drone-based 2D measurements of cross-polar beam structure and of polarization alignment of an array. The high frequency and spatial resolution achieved with this system have revealed the rich structure of the beam of each antenna, and enabled comparisons between individual dishes and to electromagnetic simulations., Comment: Submitted to ApJ

Reproducibility and Open Science are increasingly discussed as essential aspects of the research process. While there are areas where the Lattice community has been ahead of the curve with respect to the broader research world in this space, including early adoption of open publications via the arXiv, and the introduction of the International Lattice Data Grid in the 2000s, there are other areas where lattice practitioners could benefit from practices already adopted in other disciplines. In this Contribution, we report the outcomes of a panel discussion on this topic at the Lattice 2024 conference; after a discussion on motivations for work in this space, and introductory discussions of the relevant experiences of the panelists, we provide summaries of answers to the questions posed by the audience in the panel., Comment: 18 pages, 1 figure. Proceedings of the 41st International Symposium on Lattice Field Theory (Lattice 2024), July 28th - August 3rd, 2024, University of Liverpool, UK

This work introduces SpinGlassPEPS$.$jl, a software package implemented in Julia, designed to find low-energy configurations of generalized Potts models, including Ising and QUBO problems, utilizing heuristic tensor network contraction algorithms on quasi-2D geometries. In particular, the package employs the Projected Entangled-Pairs States to approximate the Boltzmann distribution corresponding to the model's cost function. This enables an efficient branch-and-bound search (within the probability space) that exploits the locality of the underlying problem's topology. As a result, our software enables the discovery of low-energy configurations for problems on quasi-2D graphs, particularly those relevant to modern quantum annealing devices. The modular architecture of SpinGlassPEPS$.$jl supports various contraction schemes and hardware acceleration.

Due to its state-of-the-art estimation performance complemented by rigorous and non-conservative uncertainty bounds, Gaussian process regression is a popular tool for enhancing dynamical system models and coping with their inaccuracies. This has enabled a plethora of successful implementations of Gaussian process-based model predictive control in a variety of applications over the last years. However, despite its evident practical effectiveness, there are still many open questions when attempting to analyze the associated optimal control problem theoretically and to exploit the full potential of Gaussian process regression in view of safe learning-based control. The contribution of this review is twofold. The first is to survey the available literature on the topic, highlighting the major theoretical challenges such as (i) addressing scalability issues of Gaussian process regression; (ii) taking into account the necessary approximations to obtain a tractable MPC formulation; (iii) including online model updates to refine the dynamics description, exploiting data collected during operation. The second is to provide an extensive discussion of future research directions, collecting results on uncertainty quantification that are related to (but yet unexploited in) optimal control, among others. Ultimately, this paper provides a toolkit to study and advance Gaussian process-based model predictive control.

Density ratio estimation in high dimensions can be reframed as integrating a certain quantity, the time score, over probability paths which interpolate between the two densities. In practice, the time score has to be estimated based on samples from the two densities. However, existing methods for this problem remain computationally expensive and can yield inaccurate estimates. Inspired by recent advances in generative modeling, we introduce a novel framework for time score estimation, based on a conditioning variable. Choosing the conditioning variable judiciously enables a closed-form objective function. We demonstrate that, compared to previous approaches, our approach results in faster learning of the time score and competitive or better estimation accuracies of the density ratio on challenging tasks. Furthermore, we establish theoretical guarantees on the error of the estimated density ratio.

Large language models, employed as multiple agents that interact and collaborate with each other, have excelled at solving complex tasks. The agents are programmed with prompts that declare their functionality, along with the topologies that orchestrate interactions across agents. Designing prompts and topologies for multi-agent systems (MAS) is inherently complex. To automate the entire design process, we first conduct an in-depth analysis of the design space aiming to understand the factors behind building effective MAS. We reveal that prompts together with topologies play critical roles in enabling more effective MAS design. Based on the insights, we propose Multi-Agent System Search (MASS), a MAS optimization framework that efficiently exploits the complex MAS design space by interleaving its optimization stages, from local to global, from prompts to topologies, over three stages: 1) block-level (local) prompt optimization; 2) workflow topology optimization; 3) workflow-level (global) prompt optimization, where each stage is conditioned on the iteratively optimized prompts/topologies from former stages. We show that MASS-optimized multi-agent systems outperform a spectrum of existing alternatives by a substantial margin. Based on the MASS-found systems, we finally propose design principles behind building effective multi-agent systems., Comment: 11 pages, 7 figures, 1 table (30 pages, 9 figures, 5 tables including references and appendices)

The article considers the Derrida-Retaux model with a random number of terms, i.e. a sequence of integer random variables defined by the relations $ X_{n + 1} = (X_n^{(1)} + X_n^{(2)} + ... + X_n^{(N_n)} - a)^{+}$, $n\ge 0$, where $X_n^{j}$ are independent copies of $X_n$, the values of $N_j$ are independent and identically distributed, $a$ is a positive integer. The energy in the model is defined as $Q:=\lim\limits_{n\to\infty} \frac{\mathbb{E}(X_{n})}{(\mathbb{E}N_1)^{n}}$. We present sufficient conditions (in terms of distributions of $X_0$ and $N_1$) for subcritical ($Q=0$) and supercritical ($Q>0$) regimes of model behavior.

We study fair vertex problem metatheorems on graphs within the scope of structural parameterization in parameterized complexity. Unlike typical graph problems that seek the smallest set of vertices satisfying certain properties, the goal here is to find such a set that does not contain too many vertices in any neighborhood of any vertex. Formally, the task is to find a set $X\subseteq V(G)$ of fair cost $k$, i.e., such that for all $v\in V(G)$ $|X\cap N(v)|\le k$. Recently, Knop, Masa\v{r}\'ik, and Toufar [MFCS 2019] showed that all fair MSO$_1$ definable problems can be solved in FPT time parameterized by the twin cover of a graph. They asked whether such a statement would be achievable for a more general parameterization of cluster vertex deletion, i.e., the smallest number of vertices required to be removed from the graph such that what remains is a collection of cliques. In this paper, we prove that in full generality this is not possible by demonstrating a W[1]-hardness. On the other hand, we give a sufficient property under which a fair MSO$_1$ definable problem admits an FPT algorithm parameterized by the cluster vertex deletion number. Our algorithmic formulation is very general as it captures the fair variant of many natural vertex problems such as the Fair Feedback Vertex Set, the Fair Vertex Cover, the Fair Dominating Set, the Fair Odd Cycle Transversal, as well as a connected variant of thereof. Moreover, we solve the Fair $[\sigma,\rho]$-Domination problem for $\sigma$ finite, or $\sigma=\mathbb{N}$ and $\rho$ cofinite. Specifically, given finite or cofinite $\rho\subseteq \mathbb{N}$ and finite $\sigma$, or $\rho\subseteq \mathbb{N}$ cofinite and $\sigma=\mathbb{N}$, the task is to find set of vertices $X\subseteq V(G)$ of fair cost at most $k$ such that for all $v\in X$, $|N(v)\cap X|\in\sigma$ and for all $v\in V(G)\setminus X$, $|N(v)\cap X|\in\rho$., Comment: 23 pages, 1 figure

Conversational assistants process personal data and must comply with data protection regulations that require providers to be transparent with users about how their data is handled. Transparency, in a legal sense, demands preciseness, comprehensibility and accessibility, yet existing solutions fail to meet these requirements. To address this, we introduce a new human-expert-generated dataset for Privacy Question-Answering (Q&A), developed through an iterative process involving legal professionals and conversational designers. We evaluate this dataset through linguistic analysis and a user study, comparing it to privacy policy excerpts and state-of-the-art responses from Amazon Alexa. Our findings show that the proposed answers improve usability and clarity compared to existing solutions while achieving legal preciseness, thereby enhancing the accessibility of data processing information for Conversational AI and Natural Language Processing applications., Comment: Submitted to CHI'25

Nonlinear elastic properties of polymers and polymeric composites are essential for accurate prediction of their response to dynamic loads, which is crucial in a wide range of applications. These properties can be affected by strain rate, temperature, and pressure. The temperature susceptibility of nonlinear elastic moduli of polymers remains poorly understood. We have recently observed a significant frequency dependence of the nonlinear elastic (Murnaghan) moduli of polystyrene. In this paper we expand this analysis by the temperature dependence. The measurement methodology was based on the acousto-elastic effect, and involved analysis of the dependencies of velocities of longitudinal and shear single-frequency ultrasonic waves in the sample on the applied static pressure. Measurements were performed at different temperatures in the range of 25-65 {\deg}C and at different frequencies in the range of 0.75-3 MHz. The temperature susceptibility of the nonlinear moduli $l$ and $m$ was found to be two orders of magnitude larger than that of linear moduli $\lambda$ and $\mu$. At the same time, the observed variations of $n$ modulus with temperature were low and within the measurement tolerance. The observed tendencies can be explained by different influence of pressure on relaxation processes in the material at different temperatures.

Two-stage exams, which pair a traditional individual exam with a subsequent collaborative exam, are widely popular with both students and faculty for fostering deep engagement, collaboration, and immediate feedback. Over the last decade, this assessment model has gained substantial traction in STEM college courses; however, holding both stages during class sessions often limits the full potential of two-stage exams, deterring many instructors from adopting them. To accommodate these constraints, instructors must either reduce the exam's content depth and coverage, or hold evening, in-person exam sessions at fixed times, which can often be impractical for students and faculty with external commitments. In this paper, we introduce an alternative asynchronous approach that allows the collaborative portion to be completed unsupervised and outside of regular class hours, within 48 hours of the individual exam. This method not only eases logistical constraints but also allows more time for collaboration, potentially enhancing feedback quality. Our findings show that students equate their engagement, collaboration, and feedback quality during the asynchronous collaborative exam with that of standard in-class exams and show signs of retention of learning. This cost-effective approach requires no extra time or resources and could promote widespread adoption of this effective assessment method, especially in online-only and general STEM courses., Comment: 5 figure, 1 table

Standard reinforcement learning (RL) assumes that an agent can observe a reward for each state-action pair. However, in practical applications, it is often difficult and costly to collect a reward for each state-action pair. While there have been several works considering RL with trajectory feedback, it is unclear if trajectory feedback is inefficient for learning when trajectories are long. In this work, we consider a model named RL with segment feedback, which offers a general paradigm filling the gap between per-state-action feedback and trajectory feedback. In this model, we consider an episodic Markov decision process (MDP), where each episode is divided into $m$ segments, and the agent observes reward feedback only at the end of each segment. Under this model, we study two popular feedback settings: binary feedback and sum feedback, where the agent observes a binary outcome and a reward sum according to the underlying reward function, respectively. To investigate the impact of the number of segments $m$ on learning performance, we design efficient algorithms and establish regret upper and lower bounds for both feedback settings. Our theoretical and experimental results show that: under binary feedback, increasing the number of segments $m$ decreases the regret at an exponential rate; in contrast, surprisingly, under sum feedback, increasing $m$ does not reduce the regret significantly.

Health information technologies are transforming how mental healthcare is paid for through value-based care programs, which tie payment to data quantifying care outcomes. But, it is unclear what outcomes data these technologies should store, how to engage users in data collection, and how outcomes data can improve care. Given these challenges, we conducted interviews with 30 U.S.-based mental health clinicians to explore the design space of health information technologies that support outcomes data specification, collection, and use in value-based mental healthcare. Our findings center clinicians' perspectives on aligning outcomes data for payment programs and care; opportunities for health technologies and personal devices to improve data collection; and considerations for using outcomes data to hold stakeholders including clinicians, health insurers, and social services financially accountable in value-based mental healthcare. We conclude with implications for future research designing and developing technologies supporting value-based care across stakeholders involved with mental health service delivery.

Correlation among the observations in high-dimensional regression modeling can be a major source of confounding. We present a new open-source package, plmmr, to implement penalized linear mixed models in R. This R package estimates correlation among observations in high-dimensional data and uses those estimates to improve prediction with the best linear unbiased predictor. The package uses memory-mapping so that genome-scale data can be analyzed on ordinary machines even if the size of data exceeds RAM. We present here the methods, workflow, and file-backing approach upon which plmmr is built, and we demonstrate its computational capabilities with two examples from real GWAS data., Comment: 23 pages, 5 figures; https://github.com/pbreheny/plmmr

V498 Hya (SDSS J084555.07+033929.2) was identified as a short-period cataclysmic variable (CV) by the Catalina Real-Time Transient Survey (CRTS) in 2008. The superhump period was measured during the detected single superoutburst of V498 Hya. The quiescent spectrum subsequently taken by the \SDSSV\ Milky Way Mapper survey suggested that the CV donor may be a brown dwarf. We present time-resolved follow-up spectroscopy of V498 Hya in quiescence, obtained with the GTC OSIRIS spectrograph, from which we derived the 86.053 min spectroscopic period, systemic radial velocity, and the gravitational redshift of the Mg II line. We also modeled the spectral energy distribution to constrain the system parameters, including the > 0.82 Ms mass of the white dwarf and the best-fit value 0.043 +/- 0.004 Ms of the donor star mass. This combination of parameters implies that V498 Hya has evolved past the period minimum and is a relatively rare ``period bouncer''., Comment: The manuscript has been accepted for publication in Monthly Notices of the Royal Astronomical Society Main Journal. Accepted 2025 January 30. Received 2025 January 30; in original form 2024 June 25

String-net models describe a vast family of topological orders in two spatial dimensions, but fail to produce all the expected anyonic excitations. Following arXiv:1502.03433, we consider an extended string-net model by attaching one tail to each plaquette of the lattice, allowing all anyons to emerge as elementary plaquette excitations for arbitrary input categories. The corresponding tube algebra is the mathematical tool needed to construct the anyons from the input category and to obtain their internal multiplicities. We use them to compute the energy level degeneracies and the partition function. In the thermodynamic limit, the latter is dominated by the trivial (vacuum) anyon, so that the topological order is destroyed at any non-zero temperature. In a finite-size system, order survives up to a finite temperature, similarly to the one-dimensional classical Ising model. We confirm this by computing thermal averages of topological projectors, Wegner-Wilson loops and the topological mutual information. The results are also generalized to models with multiple tails per plaquette., Comment: 19 pages, 6 figures

Prokaryotes have evolved various mechanisms to counter viruses, which in their turn developed numerous strategies to avoid defenses of the hosts. Dozens of such defense and counter-defense mechanisms have recently been discovered, yet the number of such systems held by a given virus or its host is limited. Here, we present numerical and theoretical arguments for the existence of the maximal number of ecologically and evolutionary sustainable defense and counter-defense systems maintained by both sides at any time of the never-ending evolutionary arms race. We find that the number of such systems is of the order of 10 for a broad range of assumptions about the costs and benefits of defense and counter-defense mechanisms and their specificity. This optimum in the number of defense and counter-defense systems appears as a result of a compromise between the metabolic and autoimmune costs of adding a new layer of defense and the benefits it conveys., Comment: 16 pages, 10 figures

The recent success of specialized Large Language Models (LLMs) in domains such as mathematical reasoning and coding has led to growing interest in methods for merging these expert LLMs into a unified Mixture-of-Experts (MoE) model, with the goal of enhancing performance in each domain while retaining effectiveness on general tasks. However, the effective merging of expert models remains an open challenge, especially for models with highly divergent weight parameters or different architectures. State-of-the-art MoE merging methods only work with homogeneous model architectures and rely on simple unweighted averaging to merge expert layers, which does not address parameter interference and requires extensive fine-tuning of the merged MoE to restore performance. To address these limitations, this paper introduces new MoE merging techniques, including strategies to mitigate parameter interference, routing heuristics to reduce the need for MoE fine-tuning, and a novel method for merging experts with different architectures. Extensive experiments across multiple domains demonstrate the effectiveness of our proposed methods, reducing fine-tuning costs, improving performance over state-of-the-art methods, and expanding the applicability of MoE merging., Comment: Accepted by NAACL 2025 Main