Search

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: 25 pages, 6 figures

Continuous gravitational waves (CWs) emission from neutron stars carries information about their internal structure and equation of state, and it can provide tests of General Relativity. We present a search for CWs from a set of 45 known pulsars in the first part of the fourth LIGO--Virgo--KAGRA observing run, known as O4a. We conducted a targeted search for each pulsar using three independent analysis methods considering the single-harmonic and the dual-harmonic emission models. We find no evidence of a CW signal in O4a data for both models and set upper limits on the signal amplitude and on the ellipticity, which quantifies the asymmetry in the neutron star mass distribution. For the single-harmonic emission model, 29 targets have the upper limit on the amplitude below the theoretical spin-down limit. The lowest upper limit on the amplitude is $6.4\!\times\!10^{-27}$ for the young energetic pulsar J0537-6910, while the lowest constraint on the ellipticity is $8.8\!\times\!10^{-9}$ for the bright nearby millisecond pulsar J0437-4715. Additionally, for a subset of 16 targets we performed a narrowband search that is more robust regarding the emission model, with no evidence of a signal. We also found no evidence of non-standard polarizations as predicted by the Brans-Dicke theory., Comment: main paper: 12 pages, 6 figures, 4 tables

In order to assist higher education institutions and their students during the pandemic, the federal government established the Higher Education Emergency Relief (HEER) Fund, which directed over $75 billion to institutions of higher education--including nearly $25 billion to community colleges--over a three-year period. The U.S. Department of Education worked on a rapid timeline to distribute these funds to institutions, which they could use to provide direct aid to students facing financial challenges and cover institutional costs related to the pandemic. Drawing on a survey of community colleges in six states--California, Michigan, New York (State University of New York [SUNY] colleges), Ohio, Tennessee, and Texas--this report provides insight into the specific pandemic recovery activities colleges implemented, colleges' perceptions of how successful HEER funds were in addressing student and institutional needs during the pandemic, and colleges' views of unmet needs. The institutional survey was completed by 170 out of a total of 265 community colleges in the six states. Key findings from the report: (1) Colleges spent nearly all the HEER funds they received. Given the large amount of HEER funding and the fact that colleges did not need to submit a proposal and budget for how they would use the funds, it should not be assumed that colleges would have spent all the money they received. Yet colleges spent nearly all the funds they received by the time the HEER program ended in June 2023; (2) HEER funds met a variety of student and institutional needs during the pandemic. Colleges had relatively few problems using the funds and felt that the aid was successful in mitigating student and institutional hardships; (3) Colleges focused on retaining existing students; they employed a variety of methods to support students in need. Colleges used HEER funds to support and retain existing (pre-pandemic) students rather than to recruit new students. They focused on supporting students with financial exigencies, including those experiencing food and housing insecurity. They used institutional aid to forgive debt owed to the college and to provide food, housing, and childcare assistance; (4) Spending patterns suggest that colleges experienced similar challenges during the pandemic and often prioritized the same objectives. Despite differences in state contexts and institutional settings, colleges tended to allocate funds in similar ways. For example, most colleges used aid for campus safety and technology hardware. Expenditure patterns also shifted over time in similar ways, indicating that colleges were responsive to evolving needs; (5) Expenditures related to campus safety and technology remained strong but decreased in frequency over time; expenditures to support students' mental health increased in frequency. Mental health services was the only expenditure category that increased in frequency in each of the three years of funding, likely reflecting the toll the pandemic took on students' mental health; (6) Comparing pre- and post-pandemic spending, HEER funds had the most impact on increasing support for technology hardware, high-speed internet, and housing assistance. Colleges used HEER funds both to fund existing services and to begin offering new ones. Fewer than a third of colleges had services in place to provide technology hardware, high-speed internet, and housing assistance before the pandemic; many more did so afterward; (7) Concerns about the end of HEER funding and priorities for future funding expose a need for continued flexible resources to address students' financial needs. Colleges' main concern about the end of HEER funding was that it would limit their ability to support students during an emergency. Their top priority for using future funding was additional student aid; and (8) Rural and vocational/technical colleges (as defined by the Carnegie Classification) may have had fewer resources prior to the pandemic and may be in greater need of additional support. Colleges in towns and rural areas and colleges focused on technical training were less likely to offer a number of supports both pre- and post-pandemic. Rural colleges were also less likely to report having received additional funding for pandemic recovery from sources other than HEER funds. Overall, while the survey findings suggest that HEER largely met the goals for which it was intended, they also point to the importance of addressing systemic challenges facing community college students and the institutions that serve them. Now that the immediate crisis of the pandemic has passed and HEER funding has ended, there is an opportunity to think strategically about the investments that are needed to promote student success over the long term, particularly for underserved and financially vulnerable students who are the most at risk of stopping out or not enrolling in the first place.

Active, responsive, nonequilibrium materials, at the forefront of materials engineering, offer dynamical restructuring, mobility and other complex life-like properties. Yet, this enhanced functionality comes with significant amplification of the size and complexity of the datasets needed to characterize their properties, thereby challenging conventional approaches to analysis. To meet this need, we present BARCODE (Biomaterial Activity Readouts to Categorize, Optimize, Design and Engineer), an open-access software that automates high throughput screening of microscopy video data to enable nonequilibrium material optimization and discovery. BARCODE produces a unique fingerprint or barcode of performance metrics that visually and quantitatively encodes dynamic material properties with minimal file size. Using three complementary material agnostic analysis branches, BARCODE significantly reduces data dimensionality and size, while providing rich, multiparametric outputs and rapid tractable characterization of activity and structure. We analyze a series of datasets of cytoskeleton networks and cell monolayers to demonstrate the ability of BARCODE to accelerate and streamline screening and analysis, reveal unexpected correlations and emergence, and enable broad non-expert data access, comparison, and sharing.

We present a structural analysis of the young massive star cluster Westerlund 1 (Wd 1). With multi-epoch Hubble Space Telescope (HST) observations, we measure the proper motions of $10346$ stars and determine their kinematic memberships by fitting a Gaussian mixture model to their proper motions. After correcting for extinction and completeness, we model the stellar density distribution and confirm the presence of an elongation with an eccentricity of $0.71$. The eccentricity decreases slightly with increasing mass. We fit the radial profile with the Elson, Fall, and Freeman model, observing a decrease in the core radius with increasing mass, indicative of weak but detectable mass segregation. This finding is further supported by a measured mass segregation ratio of $\Lambda_\mathrm{\rm MSR}=1.11\pm0.11$, only above $1$ by $1\sigma$, and slightly shorter minimum spanning tree length for higher mass bins. The cluster has a 1D velocity dispersion of $3.42 \pm 0.10~\mathrm{km}\,\mathrm{s}^{-1}$, suggesting it is subvirial. The subvirial state implies either exceptionally high star formation efficiency or inefficient stellar feedback caused by local gas expulsion before stars reach the cluster. The crossing time is $0.30$ Myr and the relaxation time is $0.26$ Gyr. Given the age of Wd 1 of $10.7$ Myr, we expect evident mass segregation for stars more massive than $10~M_\odot$, which accounts for the minor mass segregation found in the mass range of $1.00\unicode{x2013}12.14~M_\odot$ in this work. This suggests the overall mass segregation in Wd 1 is not primordial., Comment: 26 pages, 22 figures, 6 tables

Automatic sleep staging based on electroencephalography (EEG) and electromyography (EMG) signals is an important aspect of sleep-related research. Current sleep staging methods suffer from two major drawbacks. First, there are limited information interactions between modalities in the existing methods. Second, current methods do not develop unified models that can handle different sources of input. To address these issues, we propose a novel sleep stage scoring model sDREAMER, which emphasizes cross-modality interaction and per-channel performance. Specifically, we develop a mixture-of-modality-expert (MoME) model with three pathways for EEG, EMG, and mixed signals with partially shared weights. We further propose a self-distillation training scheme for further information interaction across modalities. Our model is trained with multi-channel inputs and can make classifications on either single-channel or multi-channel inputs. Experiments demonstrate that our model outperforms the existing transformer-based sleep scoring methods for multi-channel inference. For single-channel inference, our model also outperforms the transformer-based models trained with single-channel signals.

Vision language models have achieved impressive results across various fields. However, adoption in remote sensing remains limited, largely due to the scarcity of paired image-text data. To bridge this gap, synthetic caption generation has gained interest, traditionally relying on rule-based methods that use metadata or bounding boxes. While these approaches provide some description, they often lack the depth needed to capture complex wide-area scenes. Large language models (LLMs) offer a promising alternative for generating more descriptive captions, yet they can produce generic outputs and are prone to hallucination. In this paper, we propose a new method to enhance vision-language datasets for remote sensing by integrating maps as external data sources, enabling the generation of detailed, context-rich captions. Additionally, we present methods to measure and mitigate hallucinations in LLM-generated text. We introduce fMoW-mm, a multimodal dataset incorporating satellite imagery, maps, metadata, and text annotations. We demonstrate its effectiveness for automatic target recognition in few-shot settings, achieving superior performance compared to other vision-language remote sensing datasets.

In this work, we study the asymptotic behavior of Mixture of Experts (MoE) trained via gradient flow on supervised learning problems. Our main result establishes the propagation of chaos for a MoE as the number of experts diverges. We demonstrate that the corresponding empirical measure of their parameters is close to a probability measure that solves a nonlinear continuity equation, and we provide an explicit convergence rate that depends solely on the number of experts. We apply our results to a MoE generated by a quantum neural network.

Understanding the transition from atomic gas to molecular gas is critical to explain the formation and evolution of molecular clouds. However, the gas phases involved, cold HI and CO-dark molecular gas, are challenging to directly observe and physically characterize. We observed the Cygnus X star-forming complex in carbon radio recombination lines (CRRLs) at 274--399 MHz with the Green Bank Telescope at 21 pc (48') resolution. Of the 30 deg^2 surveyed, we detect line-synthesized C273alpha emission from 24 deg^2 and produce the first large-area maps of low-frequency CRRLs. The morphology of the C273alpha emission reveals arcs, ridges, and extended possibly sheet-like gas which are often on the outskirts of CO emission and likely transitioning from HI-to-H_2. The typical angular separation of C273alpha and 13CO emission is 12 pc, and we estimate C273alpha gas densities of n_H ~ 40 - 400 cm^3. The C273alpha line profiles are Gaussian and likely turbulent broadened, spanning a large range of FWHM from 2 to 20 km/s with a median of 10.6 km/s. Mach numbers fall within 10--30. The turbulent timescale is relatively short, 2.6 Myr, and we deduce that the turbulent pressure likely dominates the evolution of the C273alpha gas. Velocity offsets between C273alpha and 13CO components are apparent throughout the region and have a typical value of 2.9 km/s. Two regimes have emerged from the data: one regime in which C273alpha and 13CO are strongly related (at N_H ~ 4 x 10^21 cm^-2), and a second, in which C273alpha emits independently of the 13CO intensity. In the former regime, C273alpha may arise from the the envelopes of massive clouds (filaments), and in the latter, C273alpha emits from cold clumps in a more-diffuse mix of HI and H_2 gas., Comment: 26 pages. 13 figures. submitted to ApJ. Comments welcome!

Changing-look quasars exhibit dramatic variability in broad emission-line fluxes on short timescales. This behavior is challenging to many models of the quasar broad line region, due in large part to the short transition times between high and low states. In order to constrain the cause of the dramatic variability, we obtained contemporaneous Hubble Space Telescope UV and Hobby Eberly Telescope optical spectra of three changing-look quasars caught in their low state. We use these spectra, along with archival spectra taken during both the high and low states, to investigate potential scenarios for the change in state. Our data strongly disfavor a variable dust obscuration scenario for these three CLQs, and instead suggest that the observed transformation reflects a change in the intrinsic luminosity of the central engine. We also find that the low-state spectral energy distributions of all three quasars are reminiscent of those of low-luminosity active galactic nuclei, which suggests that the transition may result from a change in accretion flow structure caused by a reduced Eddington ratio., Comment: 20 pages, 10 figures. Accepted to ApJ January 22, 2025

From $>1000$ orbits of HST imaging, we present deep homogeneous resolved star color-magnitude diagrams that reach the oldest main sequence turnoff and uniformly measured star formation histories (SFHs) of 36 dwarf galaxies ($-6 \ge M_V \ge -17$) associated with the M31 halo, and for 10 additional fields in M31, M33, and the Giant Stellar Stream. From our SFHs we find: i) the median stellar age and quenching epoch of M31 satellites correlate with galaxy luminosity and galactocentric distance. Satellite luminosity and present-day distance from M31 predict the satellite quenching epoch to within $1.8$ Gyr at all epochs. This tight relationship highlights the fundamental connection between satellite halo mass, environmental history, and star formation duration. ii) There is no difference between the median SFH of galaxies on and off the great plane of Andromeda satellites. iii) $\sim50$\% of our M31 satellites show prominent ancient star formation ($>12$ Gyr ago) followed by delayed quenching ($8-10$ Gyr ago), which is not commonly observed among the MW satellites. iv) A comparison with TNG50 and FIRE-2 simulated satellite dwarfs around M31-like hosts show that some of these trends (dependence of SFH on satellite luminosity) are reproduced in the simulations while others (dependence of SFH on galactocentric distance, presence of the delayed-quenching population) are weaker or absent. We provide all photometric catalogs and SFHs as High-Level Science Products on MAST., Comment: Accepted for publication on ApJ. 47 pages, 24 figures, 12 tables. Corresponding HLSP data can be retrieved at: https://archive.stsci.edu/hlsp/m31-satellites

We explore covert communication of qubits over an arbitrary quantum channel. Covert communication conceals the transmissions in the channel noise, ensuring that an adversary is unable to detect their presence. We show the achievability of a $\textit{square root law}$ (SRL) for quantum covert communication similar to that for classical: $M(n)\propto\sqrt{n}$ qubits can be transmitted covertly and reliably over $n$ uses of a general quantum channel. We lower bound $M(n)$ with and without assistance from a two-way covert classical channel. In the former case, we quantify the number of classical covert bits sufficient for our protocol.

Fixed-node diffusion quantum Monte Carlo (FN-DMC) is a widely-trusted many-body method for solving the Schr\"{o}dinger equation, known for its reliable predictions of material and molecular properties. Furthermore, its excellent scalability with system complexity and near-perfect utilization of computational power makes FN-DMC ideally positioned to leverage new advances in computing to address increasingly complex scientific problems. Even though the method is widely used as a computational gold standard, reproducibility across the numerous FN-DMC code implementations has yet to be demonstrated. This difficulty stems from the diverse array of DMC algorithms and trial wave functions, compounded by the method's inherent stochastic nature. This study represents a community-wide effort to address the titular question, affirming that: Yes, FN-DMC is reproducible (when handled with care). Using the water-methane dimer as the canonical test case, we compare results from eleven different FN-DMC codes and show that the approximations to treat the non-locality of pseudopotentials are the primary source of the discrepancies between them. In particular, we demonstrate that, for the same choice of determinantal component in the trial wave function, reliable and reproducible predictions can be achieved by employing the T-move (TM), the determinant locality approximation (DLA), or the determinant T-move (DTM) schemes, while the older locality approximation (LA) leads to considerable variability in results. This work lays the foundation to establish accurate and reproducible FN-DMC estimates for all future studies across applications in materials science, physics, chemistry, and biology.

Current self-supervised learning methods for 3D medical imaging rely on simple pretext formulations and organ- or modality-specific datasets, limiting their generalizability and scalability. We present 3DINO, a cutting-edge SSL method adapted to 3D datasets, and use it to pretrain 3DINO-ViT: a general-purpose medical imaging model, on an exceptionally large, multimodal, and multi-organ dataset of ~100,000 3D medical imaging scans from over 10 organs. We validate 3DINO-ViT using extensive experiments on numerous medical imaging segmentation and classification tasks. Our results demonstrate that 3DINO-ViT generalizes across modalities and organs, including out-of-distribution tasks and datasets, outperforming state-of-the-art methods on the majority of evaluation metrics and labeled dataset sizes. Our 3DINO framework and 3DINO-ViT will be made available to enable research on 3D foundation models or further finetuning for a wide range of medical imaging applications.

We combine infrared (IR) observations collected by the James Webb Space Telescope with optical deep images by the Hubble Space Telescope taken approximately 20 years earlier to compute proper-motion membership for the globular cluster (GC) M 4 (NGC 6121) along its entire white dwarf (WD) cooling sequence (CS). These new IR observations allow us, for only the second time in a GC, to compare WD models with observations over a wide range of wavelengths, constraining fundamental astrophysical properties of WDs. Furthermore, we investigate the presence of WDs with IR excess along the WD CS of M 4, similar to the recent study conducted on the GC NGC 6397. We also determine the age difference between M 4 and NGC 6397 by comparing the absolute F150W2 magnitudes of the luminosity function peak at the bottom of the observed WD CS, and find that M 4 is slightly younger, by 0.8+/-0.5 Gyr., Comment: 13 pages, 7 figures. Accepted for publication on Astronomische Nachrichten on 15th Jan 2025. Manuscript ID: ASNA.20240125

We investigate the profitability and risk of energy storage arbitrage in electricity markets under price uncertainty, exploring both robust and chance-constrained optimization approaches. We analyze various uncertainty representations, including polyhedral, ellipsoidal uncertainty sets and probabilistic approximations, to model price fluctuations and construct efficient frontiers that highlight the tradeoff between risk and profit. Using historical electricity price data, we quantify the impact of uncertainty on arbitrage strategies and compare their performance under distinct market conditions. The results reveal that arbitrage strategies under uncertainties can effectively secure expected profits, and robust strategies perform better in risk management across varying levels of conservativeness, especially under highly volatile market conditions. This work provides insights into storage arbitrage strategy selection for market participants with differing risk preferences, emphasizing the adaptability of efficient frontiers to the electricity market.

Incorporating cells within active biomaterial scaffolds is a promising strategy to develop forefront materials that can autonomously sense, respond, and alter the scaffold in response to environmental cues or internal cell circuitry. Using dynamic biocompatible scaffolds that can self-alter their properties via crosslinking and motor-driven force-generation opens even greater avenues for actuation and control. However, the design principles associated with engineering active scaffolds embedded with cells are not well established. To address this challenge, we design a dynamic scaffold material of bacteria cells embedded within a composite cytoskeletal network of actin and microtubules that can be passively or actively crosslinked by either biotin-streptavidin or multimeric kinesin motors. Using quantitative microscopy, we demonstrate the ability to embed cells of volume fractions 0.4 to 2% throughout the network without compromising the structural integrity of the network or inhibiting crosslinking or motor-driven dynamics. Our findings suggest that both passive and active crosslinking promote entrainment of cells within the network, while depletion interactions play a more important role in uncrosslinked networks. Moreover, we show that large-scale structures emerge with the addition of cell fractions as low as 0.4%, but these structures do not influence the microscale structural lengthscale of the materials. Our work highlights the potential of our composite biomaterial in designing autonomous materials controlled by cells, and provides a roadmap for effectively coupling cells to complex composite materials with an eye towards using cells as in situ factories to program material modifications.

The cytoskeleton is an active composite of filamentous proteins that dictates diverse mechanical properties and processes in eukaryotic cells by generating forces and autonomously restructuring itself. Enzymatic motors that act on the comprising filaments play key roles in this activity, driving spatiotemporally heterogeneous mechanical responses that are critical to cellular multifunctionality, but also render mechanical characterization challenging. Here, we couple optical tweezers microrheology and fluorescence microscopy with simulations and mathematical modeling to robustly characterize the mechanics of active composites of actin filaments and microtubules restructured by kinesin motors. We discover that composites exhibit a rich ensemble of force response behaviors, elastic, yielding, and stiffening, with their propensity and properties tuned by motor concentration and strain rate. Moreover, intermediate kinesin concentrations elicit emergent mechanical stiffness and resistance while higher and lower concentrations exhibit softer, more viscous dissipation. We further show that composites transition from well-mixed interpenetrating double-networks of actin and microtubules to de-mixed states of microtubule-rich aggregates surrounded by relatively undisturbed actin phases. It is this de-mixing that leads to the emergent mechanical response, offering an alternate route that composites can leverage to achieve enhanced stiffness through coupling of structure and mechanics.

In this paper, we give theoretically and practically efficient implementations of Big Atomics, i.e., $k$-word linearizable registers that support the load, store, and compare-and-swap (CAS) operations. While modern hardware supports $k = 1$ and sometimes $k = 2$ (e.g., double-width compare-and-swap in x86), our implementations support arbitrary $k$. Big Atomics are useful in many applications, including atomic manipulation of tuples, version lists, and implementing load-linked/store-conditional (LL/SC). We design fast, lock-free implementations of big atomics based on a novel fast-path-slow-path approach we develop. We then use them to develop an efficient concurrent hash table, as evidence of their utility. We experimentally validate the approach by comparing a variety of implementations of big atomics under a variety of workloads (thread counts, load/store ratios, contention, oversubscription, and number of atomics). The experiments compare two of our lock-free variants with C++ std::atomic, a lock-based version, a version using sequence locks, and an indirect version. The results show that our approach is close to the fastest under all conditions and far outperforms others under oversubscription. We also compare our big atomics based concurrent hash table to a variety of other state-of-the-art hash tables that support arbitrary length keys and values, including implementations from Intel's TBB, Facebook's Folly, libcuckoo, and a recent release from Boost. The results show that our approach of using big atomics in the design of hash tables is a promising direction.

Using the full four-year SPTpol 500 deg$^2$ dataset in both the 95 GHz and 150 GHz frequency bands, we present measurements of the temperature and $E$-mode polarization of the cosmic microwave background (CMB), as well as the $E$-mode polarization auto-power spectrum ($EE$) and temperature-$E$-mode cross-power spectrum ($TE$) in the angular multipole range $50<\ell<8000$. We find the SPTpol dataset to be self-consistent, passing several internal consistency tests based on maps, frequency bands, bandpowers, and cosmological parameters. The full SPTpol dataset is well-fit by the $\Lambda CDM$ model, for which we find $H_0=70.48\pm2.16$ km s$^{-1}$ Mpc$^{-1}$ and $\Omega_m=0.271\pm0.026$, when using only the SPTpol data and a Planck-based prior on the optical depth to reionization. The $\Lambda CDM$ parameter constraints are consistent across the 95 GHz-only, 150 GHz-only, $TE$-only, and $EE$-only data splits. Between the $\ell<1000$ and $\ell>1000$ data splits, the $\Lambda CDM$ parameter constraints are borderline consistent at the $\sim2\sigma$ level. This consistency improves when including a parameter $A_L$, the degree of lensing of the CMB inferred from the smearing of acoustic peaks. When marginalized over $A_L$, the $\Lambda CDM$ parameter constraints from SPTpol are consistent with those from Planck. The power spectra presented here are the most sensitive measurements of the lensed CMB damping tail to date for roughly $\ell > 1700$ in $TE$ and $\ell > 2000$ in $EE$.

Silicon processing techniques such as atomic precision advanced manufacturing (APAM) and epitaxial growth require surface preparations that activate oxide desorption (typically >1000 $^{\circ}$C) and promote surface reconstruction toward atomically-clean, flat, and ordered Si(100)-2$\times$1. We compare aqueous and vapor phase cleaning of Si and Si/SiGe surfaces to prepare APAM-ready and epitaxy-ready surfaces at lower temperatures. Angle resolved X-ray photoelectron spectroscopy (ARXPS) and Fourier transform infrared (FTIR) spectroscopy indicate that vapor hydrogen fluoride (VHF) cleans dramatically reduce carbon surface contamination and allow the chemically prepared surface to reconstruct at lower temperatures, 600 $^{\circ}$C for Si and 580 $^{\circ}$C for a Si/Si$_{0.7}$Ge$_{0.3}$ heterostructures, into an ordered atomic terrace structure indicated by scanning tunneling microscopy (STM). After thermal treatment and vacuum hydrogen termination, we demonstrate STM hydrogen desorption lithography (HDL) on VHF-treated Si samples, creating reactive zones that enable area-selective chemistry using a thermal budget similar to CMOS process flows. We anticipate these results will establish new pathways to integrate APAM with Si foundry processing.

Fisheries scientists use regression models to estimate population quantities, such as biomass or abundance, for use in climate, habitat, stock, and ecosystem assessments. However, these models are sensitive to the chosen probability distribution used to characterize observation error. Here, we introduce the generalized gamma distribution (GGD), which has not been widely used in fisheries science. The GGD has useful properties: (1) it reduces to the lognormal distribution when the shape parameter approaches zero; (2) it reduces to the gamma distribution when the shape and scale parameters are equal; and (3) the coefficient of variation is independent of the mean. We assess the relative performance and robustness of the GGD to estimate biomass density across different observation error types in a simulation experiment. When fit to data generated from the GGD, lognormal, gamma, and Tweedie families, the GGD had low bias and high predictive accuracy. Finally, we fit spatiotemporal index standardization models using the R package sdmTMB to 15 species from three trawl surveys from the Gulf of Alaska and coast of British Columbia, Canada. When the Akaike information criterion (AIC) weight was compared among fits using the lognormal, gamma, and Tweedie families the GGD was the most commonly selected model.

This work presents FreeSVC, a promising multilingual singing voice conversion approach that leverages an enhanced VITS model with Speaker-invariant Clustering (SPIN) for better content representation and the State-of-the-Art (SOTA) speaker encoder ECAPA2. FreeSVC incorporates trainable language embeddings to handle multiple languages and employs an advanced speaker encoder to disentangle speaker characteristics from linguistic content. Designed for zero-shot learning, FreeSVC enables cross-lingual singing voice conversion without extensive language-specific training. We demonstrate that a multilingual content extractor is crucial for optimal cross-language conversion. Our source code and models are publicly available.

At the risk of overstating the case, connectionist approaches to machine learning, i.e. neural networks, are enjoying a small vogue right now. However, these methods require large volumes of data and produce models that are uninterpretable to humans. An alternative framework that is compatible with neural networks and gradient-based learning, but explicitly models compositionality, is Vector Symbolic Architectures (VSAs). VSAs are a family of algebras on high-dimensional vector representations. They arose in cognitive science from the need to unify neural processing and the kind of symbolic reasoning that humans perform. While machine learning methods have benefited from category theoretical analyses, VSAs have not yet received similar treatment. In this paper, we present a first attempt at applying category theory to VSAs. Specifically, we conduct a brief literature survey demonstrating the lacking intersection of these two topics, provide a list of desiderata for VSAs, and propose that VSAs may be understood as a (division) rig in a category enriched over a monoid in Met (the category of Lawvere metric spaces). This final contribution suggests that VSAs may be generalised beyond current implementations. It is our hope that grounding VSAs in category theory will lead to more rigorous connections with other research, both within and beyond, learning and cognition., Comment: 13 pages, no figures, 2 tables, one appendix

Spin states in semiconductors provide exceptionally stable and noise-resistant environments for qubits, positioning them as optimal candidates for reliable quantum computing technologies. The proposal to use nuclear and electronic spins of donor atoms in silicon, introduced by Kane in 1998, sparked a new research field focused on the precise positioning of individual impurity atoms for quantum devices, utilising scanning tunnelling microscopy and ion implantation. This roadmap article reviews the advancements in the 25 years since Kane's proposal, the current challenges, and the future directions in atomic-scale semiconductor device fabrication and measurement. It covers the quest to create a silicon-based quantum computer and expands to include diverse material systems and fabrication techniques, highlighting the potential for a broad range of semiconductor quantum technological applications. Key developments include phosphorus in silicon devices such as single-atom transistors, arrayed few-donor devices, one- and two-qubit gates, three-dimensional architectures, and the development of a toolbox for future quantum integrated circuits. The roadmap also explores new impurity species like arsenic and antimony for enhanced scalability and higher-dimensional spin systems, new chemistry for dopant precursors and lithographic resists, and the potential for germanium-based devices. Emerging methods, such as photon-based lithography and electron beam manipulation, are discussed for their disruptive potential. This roadmap charts the path toward scalable quantum computing and advanced semiconductor quantum technologies, emphasising the critical intersections of experiment, technological development, and theory., Comment: 94 pages

Using JWST Near Infrared Camera (NIRCam) images of the globular cluster 47 Tucanae (or NGC 104), taken at two epochs just 7 months apart, we derived proper-motion membership down to $m_{\rm F322W2} \sim 27$. We identified an intriguing feature at the very low-mass end of the main sequence, around $\sim$ 0.08 solar masses, at magnitudes $m_{\rm F322W2} \sim 24$ and $m_{\rm F150W2} \sim 25$. This feature, dubbed "kink", is characterized by a prominent discontinuity in the slope of the main sequence. A similar discontinuity is seen in theoretical isochrones with oxygen-poor chemistries, related to the rapid onset of CH$_4$ absorption. We therefore hypothesize that the cluster hosts disproportionately more oxygen-poor stars near the bottom of the main sequence compared to the upper main sequence and the red giant branch. Our results show no strong or conclusive evidence of a rise in the brown dwarf luminosity function at faint magnitudes, in contrast to previous findings likely affected by faint red background galaxies. In our analysis, we accounted for this contamination by using proper motion membership., Comment: 13 pages, 11 figures and 1 table. Accepted for publication in Astronomy & Astrophysics on Jan 7, 2025

Processing computer vision applications (CVA) on mobile devices is challenging due to limited battery life and computing power. While cloud-based remote processing of CVA offers abundant computational resources, it introduces latency issues that can hinder real-time applications. To overcome this problem, computational offloading to edge servers has been adopted by industry and academic research. Furthermore, 5G access can also benefit CVA with lower latency and higher bandwidth than previous cellular generations. As the number of Mobile Operators and Internet Service providers relying on 5G access is growing, it is of paramount importance to elaborate a solution for supporting real time applications with the assistance of the edge computing. Besides that, open-source based platforms for Multi-access Edge Computing (MEC) and 5G core can be deployed to rapid prototyping and testing applications. This paper aims at providing an end-to-end solution of open-source MEC and 5G Core platforms along with a commercial 5G Radio. We first conceived a 5G-edge computing environment to assist near to user processing of computer vision applications. Then a sentiment analysis application is developed and integrated to the proposed 5G-Edge architecture. Finally, we conducted a performance evaluation of the proposed solution and compare it against a remote cloud-based approach in order to highlight the benefits of our proposal. The proposed architecture achieved a 260\% throughput performance increase and reduced response time by 71.3\% compared to the remote-cloud-based offloading., Comment: Accept on conference the 39th International Conference on Information Networking (ICOIN 2025): 6 pages, 8 figures, 1 table

(Abridged) While previous X-ray studies showed the dominance of regular active galactic nuclei (AGN) variability, a small fraction of sources arise from more exotic phenomena such as tidal disruption events (TDEs), quasi-periodic eruptions, or other short-lived events associated with supermassive black hole accretion. This paper describes the systematic selection of X-ray extragalactic transients found in the first two eROSITA all-sky surveys (eRASS) that are not associated with known AGN prior to eROSITA observations. We generated a variability sample from eRASS1 and eRASS2 (Dec. 2019-Dec. 2020), which includes sources with a variability significance and a fractional amplitude larger than four, located in the Legacy Survey DR10 (LS10) footprint. The properties of LS10 counterparts were used to exclude stars and known AGN. The sample was additionally cleaned using pre-eROSITA classifications, archival optical spectra, and archival X-ray data. The final catalog eRO-ExTra includes 304 extragalactic eROSITA transients and variables not associated with known AGN. More than 90% of sources have reliable LS10 optical counterparts. For each source, we provide archival X-ray data from Swift, ROSAT, and XMM-Newton; the eROSITA long-term light curve (2-2.5 years) with a light curve classification; as well as the best power law fit spectral results at the peak eROSITA epoch. Reliable spectroscopic and photometric redshifts are provided for more than 80% of the sample. Several sources in the catalog are known TDE candidates discovered by eROSITA. In addition, 31 sources are radio detected. The eRO-ExTra transients constitute a relatively clean parent sample of non-AGN variability phenomena associated with massive black holes. More than 95% of eRO-ExTra sources were discovered in X-rays with eROSITA for the first time, which makes it a valuable resource for studying unique nuclear transients., Comment: 18 pages, 15 figures, published in A&A. To download the eRO-ExTra catalog, see https://cdsarc.cds.unistra.fr/viz-bin/cat/J/A+A/693/A62

We created new reddening maps and derived new extinction laws from visual to near-infrared passbands using improved RR~Lyrae period-absolute magnitude-metallicity relations, thus enabling distance estimates for individual bulge RR~Lyrae variables. The extinction law is most uniform in RIK and RJK and the distances to individual RR~Lyrae based on these colors are determined with an accuracy six and four percent, respectively. Using only the near-infrared passbands for distance estimation we inferred the distance to the Galactic center equal to djk = 8.2 +- 0.001(stat) +- 0.53(sys)pc after geometrical correction. We show that variations in the extinction law toward the Galactic bulge can mimic a barred spatial distribution in the bulge RR~Lyrae star population in visual passbands. This arises from a gradient in extinction differences along Galactic longitudes and latitudes, which can create the perception of the Galactic bar, particularly when using visual passband-based distances. A barred angle in the RR~Lyrae spatial distribution disappears when near-infrared passband-based distances are used, as well as when reddening law variations are incorporated in visual passband-based distances. The prominence of the bar, as traced by RR~Lyrae stars, depends on their metallicity, with metal-poor RR~Lyrae stars ([Fe/H]<-1.0dex) showing little to no tilt with respect to the bar. Metal-rich ([Fe/H]>-1.0dex) RR~Lyrae stars do show a barred/bulge signature in spatial properties derived using near-infrared distances, with an angle {\iota} = 18 +- 5deg, consistent with previous bar measurements from the literature. The 5D kinematic analysis, primarily based on transverse velocities, indicates a rotational lag in RR~Lyrae stars compared to red clump giants. Despite variations in the extinction law, our kinematic conclusions are robust across different distance estimation methods., Comment: Accepted for publication in A&A

In this work, we study the weighted empirical risk minimization (weighted ERM) schema, in which an additional data-dependent weight function is incorporated when the empirical risk function is being minimized. We show that under a general ``balanceable" Bernstein condition, one can design a weighted ERM estimator to achieve superior performance in certain sub-regions over the one obtained from standard ERM, and the superiority manifests itself through a data-dependent constant term in the error bound. These sub-regions correspond to large-margin ones in classification settings and low-variance ones in heteroscedastic regression settings, respectively. Our findings are supported by evidence from synthetic data experiments., Comment: 33 pages

We perform a global QCD analysis of unpolarized parton distribution functions (PDFs) in the proton, including new $W +\,$charm production data from $pp$ collisions at the LHC, which have been suggested for constraining the strange quark PDF. In particular, we assess the impact of the $W +\,$charm data relative to that of semi-inclusive pion and kaon production data in lepton-nucleon deep-inelastic scattering. Compared with a baseline global fit that does not include these datasets, we find that semi-inclusive deep-inelastic scattering and $W +\,$charm data combined favor a larger strange distribution, but smaller than the SU(3) symmetric sea suggested in previous analyses of ATLAS data. Combined, the ratio of strange to nonstrange sea quark distributions is $R_s = (s+\bar s)/(\bar u+\bar d) \approx \{ 0.72,\, 0.48,\, 0.26 \}$ for $x = \{ 0.01,\, 0.04,\, 0.1 \}$., Comment: 33 pages, 13 figures, 3 tables

We present OLMo 2, the next generation of our fully open language models. OLMo 2 includes dense autoregressive models with improved architecture and training recipe, pretraining data mixtures, and instruction tuning recipes. Our modified model architecture and training recipe achieve both better training stability and improved per-token efficiency. Our updated pretraining data mixture introduces a new, specialized data mix called Dolmino Mix 1124, which significantly improves model capabilities across many downstream task benchmarks when introduced via late-stage curriculum training (i.e. specialized data during the annealing phase of pretraining). Finally, we incorporate best practices from T\"ulu 3 to develop OLMo 2-Instruct, focusing on permissive data and extending our final-stage reinforcement learning with verifiable rewards (RLVR). Our OLMo 2 base models sit at the Pareto frontier of performance to compute, often matching or outperforming open-weight only models like Llama 3.1 and Qwen 2.5 while using fewer FLOPs and with fully transparent training data, code, and recipe. Our fully open OLMo 2-Instruct models are competitive with or surpassing open-weight only models of comparable size, including Qwen 2.5, Llama 3.1 and Gemma 2. We release all OLMo 2 artifacts openly -- models at 7B and 13B scales, both pretrained and post-trained, including their full training data, training code and recipes, training logs and thousands of intermediate checkpoints. The final instruction model is available on the Ai2 Playground as a free research demo., Comment: Model demo available at playground.allenai.org

Our society increasingly benefits from Artificial Intelligence (AI). Unfortunately, more and more evidence shows that AI is also used for offensive purposes. Prior works have revealed various examples of use cases in which the deployment of AI can lead to violation of security and privacy objectives. No extant work, however, has been able to draw a holistic picture of the offensive potential of AI. In this SoK paper we seek to lay the ground for a systematic analysis of the heterogeneous capabilities of offensive AI. In particular we (i) account for AI risks to both humans and systems while (ii) consolidating and distilling knowledge from academic literature, expert opinions, industrial venues, as well as laypeople -- all of which being valuable sources of information on offensive AI. To enable alignment of such diverse sources of knowledge, we devise a common set of criteria reflecting essential technological factors related to offensive AI. With the help of such criteria, we systematically analyze: 95 research papers; 38 InfoSec briefings (from, e.g., BlackHat); the responses of a user study (N=549) entailing individuals with diverse backgrounds and expertise; and the opinion of 12 experts. Our contributions not only reveal concerning ways (some of which overlooked by prior work) in which AI can be offensively used today, but also represent a foothold to address this threat in the years to come., Comment: Systematization of Knowledge (SoK) paper. Accepted to the 3rd IEEE Conference on Secure and Trustworthy Machine Learning (SaTML'25)

For integral weights $\lambda$ and $\mu$ of a classical simple Lie algebra $\mathfrak{g}$, Kostant's weight multiplicity formula gives the multiplicity of the weight $\mu$ in the irreducible representation with highest weight $\lambda$, which we denote by $m(\lambda,\mu)$. Kostant's weight multiplicity formula is an alternating sum over the Weyl group of the Lie algebra whose terms are determined via a vector partition function. The Weyl alternation set $\mathcal{A}(\lambda,\mu)$ is the set of Weyl group elements that contribute nontrivially to the multiplicity $m(\lambda,\mu)$. In this article, we prove that Weyl alternation sets are order ideals in the weak Bruhat order of the corresponding Weyl group. Specializing to the Lie algebra $\mathfrak{sl}_{r+1}(\mathbb{C})$, we give a complete characterization of the Weyl alternation sets $\mathcal{A}(\tilde{\alpha},\mu)$, where $\tilde{\alpha}$ is the highest root and $\mu$ is a negative root, answering a question of Harry posed in 2024. We also provide some enumerative results that pave the way for our future work where we aim to prove Harry's conjecture that the $q$-analog of Kostant's weight multiplicity formula $m_q(\tilde{\alpha},\mu)=q^{r+j-i+1}+q^{r+j-i}-q^{j-i+1}$ when $\mu=-(\alpha_i+\alpha_{i+1}+\cdots+\alpha_{j})$ is a negative root of $\mathfrak{sl}_{r+1}(\mathbb{C})$., Comment: 24 pages, 3 figures, 2 tables

Rydberg sensors offer a unique approach to radio frequency (RF) detection, leveraging the high sensitivity and quantum properties of highly-excited atomic states to achieve performance levels beyond classical technologies. Non-linear responses and distortion behavior in Rydberg atom receivers are critical to evaluating and establishing performance metrics and capabilities such as spur-free dynamic range and tolerance to unwanted interfering signals. We report here on the measurement and characterization of non-linear behavior and spurious response of a Rydberg atomic heterodyne receiver. Single-tone and two-tone testing procedures are developed and implemented for measurement of harmonic and inter-modulation distortion in Rydberg atomic receivers based on multi-photon Rydberg spectroscopy and radio-frequency heterodyne signal detection and demodulation in an atomic vapor. For a predetermined set of atomic receiver parameters and RF carrier wave in the SHF band near-resonant to a cesium Rydberg transition, we measure and characterize atomic receiver selectivity, bandwidth, roll-off, compression point (P1dB), second-order (IP2) and third-order (IP3) intercepts, and spur-free dynamic range. Receiver intermodulation distortion is characterized for the case of an interfering signal wave applied at two frequency offsets relative to the near-resonant reference local oscillator, $\Delta F/F= 10^{-4}$ at 6dB and $10^{-6}$ at 22dB single-tone bandwidths, respectively. We observe that under suitable operating conditions the atomic receiver can exhibit a suppression of harmonic and inter-modulation distortion relative to that of classical receiver mixer amplifiers. Finally, we describe how the non-linear behaviors of atomic receivers can provide unique, controllable RF signatures inaccessible by classical counterparts and propose their use to realize secure communication modalities and applications.

As Large Language Models become integral to decision-making, optimism about their power is tempered with concern over their errors. Users may over-rely on LLM advice that is confidently stated but wrong, or under-rely due to mistrust. Reliance interventions have been developed to help users of LLMs, but they lack rigorous evaluation for appropriate reliance. We benchmark the performance of three relevant interventions by conducting a randomized online experiment with 400 participants attempting two challenging tasks: LSAT logical reasoning and image-based numerical estimation. For each question, participants first answered independently, then received LLM advice modified by one of three reliance interventions and answered the question again. Our findings indicate that while interventions reduce over-reliance, they generally fail to improve appropriate reliance. Furthermore, people became more confident after making incorrect reliance decisions in certain contexts, demonstrating poor calibration. Based on our findings, we discuss implications for designing effective reliance interventions in human-LLM collaboration.

Immediately following a disaster event, such as an earthquake, estimates of the damage extent play a key role in informing the coordination of response and recovery efforts. We develop a novel impact estimation tool that leverages a generalised Bayesian approach to generate earthquake impact estimates across three impact types: mortality, population displacement, and building damage. Inference is performed within a likelihood-free framework, and a scoring-rule-based posterior avoids information loss from non-sufficient summary statistics. We propose an adaptation of existing scoring-rule-based loss functions that accommodates the use of an approximate Bayesian computation sequential Monte Carlo (ABC-SMC) framework. The fitted model achieves results comparable to those of two leading impact estimation tools in the prediction of total mortality when tested on a set of held-out past events. The proposed method provides four advantages over existing empirical approaches: modelling produces a gridded spatial map of the estimated impact, predictions benefit from the Bayesian quantification and interpretation of uncertainty, there is direct handling of multi-shock earthquake events, and the use of a joint model between impact types allows predictions to be updated as impact observations become available., Comment: 24 pages, 21 figures and 5 tables

We present improvements to the pointing accuracy of the South Pole Telescope (SPT) using machine learning. The ability of the SPT to point accurately at the sky is limited by its structural imperfections, which are impacted by the extreme weather at the South Pole. Pointing accuracy is particularly important during SPT participation in observing campaigns with the Event Horizon Telescope (EHT), which requires stricter accuracy than typical observations with the SPT. We compile a training dataset of historical observations of astronomical sources made with the SPT-3G and EHT receivers on the SPT. We train two XGBoost models to learn a mapping from current weather conditions to two telescope drive control arguments -- one which corrects for errors in azimuth and the other for errors in elevation. Our trained models achieve root mean squared errors on withheld test data of $2.14''$ in cross-elevation and $3.57''$ in elevation, well below our goal of $5''$ along each axis. We deploy our models on the telescope control system and perform further in situ test observations during the EHT observing campaign in 2024 April. Our models result in significantly improved pointing accuracy: for sources within the range of input variables where the models are best trained, average combined pointing error improved 33%, from $15.9''$ to $10.6''$. These improvements, while significant, fall shy of our ultimate goal, but they serve as a proof of concept for the development of future models. Planned upgrades to the EHT receiver on the SPT will necessitate even stricter pointing accuracy which will be achievable with our methods., Comment: 23 pages, 9 figures, to be submitted to Journal of Astronomical Instrumentation (JAI)

Ludii is a Java general game system with a considerable number of board games, with an API for developing new agents and a game description language to create new games. To improve versatility and ease development, we provide Python interfaces for agent programming. This allows the use of Python modules to implement general game playing agents. As a means of enabling Python for creating Ludii agents, the interfaces are implemented using different Java libraries: jpy and Py4J. The main goal of this work is to determine which version is faster. To do so, we conducted a performance analysis of two different GGP algorithms, Minimax adapted to GGP and MCTS. The analysis was performed across several combinatorial games with varying depth, branching factor, and ply time. For reproducibility, we provide tutorials and repositories. Our analysis includes predictive models using regression, which suggest that jpy is faster than Py4J, however slower than a native Java Ludii agent, as expected.

Reliability analysis is a sub-field of uncertainty quantification that assesses the probability of a system performing as intended under various uncertainties. Traditionally, this analysis relies on deterministic models, where experiments are repeatable, i.e., they produce consistent outputs for a given set of inputs. However, real-world systems often exhibit stochastic behavior, leading to non-repeatable outcomes. These so-called stochastic simulators produce different outputs each time the model is run, even with fixed inputs. This paper formally introduces reliability analysis for stochastic models and addresses it by using suitable surrogate models to lower its typically high computational cost. Specifically, we focus on the recently introduced generalized lambda models and stochastic polynomial chaos expansions. These emulators are designed to learn the inherent randomness of the simulator's response and enable efficient uncertainty quantification at a much lower cost than traditional Monte Carlo simulation. We validate our methodology through three case studies. First, using an analytical function with a closed-form solution, we demonstrate that the emulators converge to the correct solution. Second, we present results obtained from the surrogates using a toy example of a simply supported beam. Finally, we apply the emulators to perform reliability analysis on a realistic wind turbine case study, where only a dataset of simulation results is available.